In the middle of May, a report, The Causes and Context of Sexual Abuse of Minors by Catholic Priests in the United States, 1950-2010^[1], was presented to the United States Conference of Catholic Bishops by researchers from John Jay College. Circulated more broadly on the 16^th of May, the document has (not without reason) been viewed as flawed to the point of being suspect.

Almost predictably from the usual quarters, the way in which the mere laity have viewed the report has been met with a near pre-reformation style of contempt. Apparently the masses need the clergy to spell it out for them!

In Australia, this high-handed misanthropy finds archetypal expression in Scott Stephens’ Catholic sexual abuse study greeted with incurious contempt^[2], in which a ranting jeremiad laments media standards on the grounds of a supposed incurious nature in relation to the Causes and Context report’s findings.

‘I suppose I should no longer be surprised by the self-righteous cynicism and seemingly wilful ignorance of the media when it comes to reporting on Catholic affairs. But it was the way that the Australian press allowed the findings of a recent study into The Causes and Context of Sexual Abuse of Minors by Catholic Priests in the United States, 1950-2010 to sail past with little more than a perfunctory acknowledgement of its existence, much less a serious engagement with its substance and implications, that has left me bristling.^[2]’ – Emphasis added.

Remember this. The standard is ‘serious engagement’ with ‘substance and implications’, and the charge is ‘self-righteous cynicism and wilful ignorance’. Don’t let these goal posts move.

Thanks to a suitably inspiring heads-up from Miranda Celeste Hale^[3], I’d spent a good deal of my free time in the days leading up to Stephens’ article reading the report, the previous Nature and Scope study^[4] (and its supplementary^[5]), appalled by what I was reading. Appalled not because it acknowledged priests had abused children (we all know that), but because the study had used a problematic definition of pedophilia, all while vaguely suggesting the moral turbulence of the 1960s and 1970s as somehow causal – please note my emphasis on vagueness.

(I could go on at length about the funding details, or the method of data collection and subsequent projections of the incidence of child abuse, but this has been addressed already at great length elsewhere by several writers).

It was on the matter of the definition of pedophilia that I sought elaboration from Stephens (or at least the ABC’s ‘Religion and Ethics’ portal), to see if certain concerns were understood, so I should first elaborate on these concerns myself.

Miranda Celeste Hale rightly points out that the definition of pedophilia used in the report deviates^[3]; that due to its terminology, sexual attraction toward and/or a sexual assault upon an eleven-to-thirteen year old, couldn’t be considered indicative of pedophilia. This variant terminology is seemingly used to reach a politically convenient conclusion.

‘It is worth noting that while the media has consistently referred to priest-abusers as “pedophile priests,” pedophilia is defined as the sexual attraction to prepubescent children. Yet, the data on priests show that 22 percent of victims were age ten and under, while the majority of victims were pubescent or postpubescent.^[1]’

The Diagnostic and Statistical Manual of Mental Disorders (DSM) defines pedophilia differently – specifically with a cut-off age of thirteen (with considerations made for the onset of pubescence in individual cases)^[6].

If you turn to page eleven of the report^[1], you will notice a graphical representation of how the Nature and Scope data has been grouped by the age groups of sexual abuse victims, and specifically, how the DSM cut-off age for pedophilia falls amidst the most populous group, rather than bounding it. The result being that according to the standard DSM definition, as Miranda points out in her post^[3], this ’22 percent of abuse victims’ line has the capacity for perpetuating gross underestimation if churned by the media uncritically.

While I agree with the churnalism concerns, I think more can be said on the matter.

Aside from the triviality of a potential public relations setback, what potential harm could have come of the report adhering more rigidly to DSM terminology? Conversely, other than potential public relations advantages, what harm could be done through the use of the report’s divergent definition?

An answer to the first question seems easy; ‘none’, which is itself quite damning, mandating considered attention by anyone undertaking serious media analysis of how the report is received.

The answer to the second question isn’t nearly as obvious.

The preventative measures recommended by the report (based on situational prevention models, education and accountability measures) focus upon limiting the opportunity for events of abuse to occur, rather than targeting the particular psychologies of child abusers. From this perspective, the question of whether priests psychologically capable of raping eleven-to-thirteen year olds, are or are not according to terminology, pedophiles, is largely a technical abstraction. An abstraction that currently doesn’t do much to guide guide policy in how to prevent the scenarios where these urges can be acted upon.

To caregivers and childcare workers who’ve undertaken anything like mandatory notification training, this approach will in general terms seem quite conventional. But limiting the discussion to this perspective, as politically convenient as it may be for religious apologetics, is irresponsible.

The history of child abuse prevention is short, and already fraught with change and revision. What if at some point in the future, a reliable psychological profile for child molesters were developed in line with the DSM definition? What if in future, such a profile could be used to filter the ordination of priests?

In any such event, the technical differences between the DSM and the Causes and Context study wouldn’t be at all trivial.

Given that any future psychological developments are more likely to be in-line with the DSM than the semantic idiosyncrasies of the Causes and Context report, an explanation of, and a justification for the difference is required, especially for us non-psychologists in the audience. Yet all we are left with in the way of justification, ostensibly, is that one definition is more flattering to reported child abusers than another.
If this ‘it’s not pedophilia if they’re over ten’ ethos gets too great a stranglehold in the Catholic Church’s attempts to deal with its own child abuse problems, or indeed, if it’s adopted as part of the standard language of religious apologetics, then I think this aspect of the terminology is worse than any potential churnalism of ’22 percent of victims’.

So what did Scott Stephens have to write on the matter?

‘Second, the study demonstrates that the overwhelming majority of victims of sexual abuse between 1950 and 2010 were male (81 per cent) and between the ages of eleven and fourteen (51 per cent). Meanwhile 27 per cent were aged fifteen to seventeen, 16 per cent were eight to ten, and only 6 percent were under the age of seven.^[2]’

While acknowledging the age groups of the victims, and clearly not diminishing the severity of the abuse, this does not address the capacity of the divergent terminology for misleading or prejudicing public discussion or policy formation.

‘Because the majority of the victims were pubescent or post-pubescent, the Causes and Context study rather controversially claims that it is not therefore strictly correct to refer to “paedophile priests” (paedophilia being defined as “the sexual attraction to prepubescent children”). While this is neither here nor there, and has no bearing on the methodology or substance of the study itself, it is a technical point that has generated unwarranted consternation among many in the media.^[2]’

Again, Stephens tells us nothing of the capacity of the study to mislead or to prejudice discussion, and even falls short of acknowledging the full definition used in the study – specifically the controversial cut-off age of ten years. You’ll understand then, when I write that I don’t think Stephens has demonstrated an understanding of the problem.

Seeking to have an understanding of the misleading nature of the definition (but not at this point its prejudicial nature) confirmed by Scott Stephens, or by the ABC’s Religion and Ethics portal, I attempted to engage them on Twitter. Stephens’ article being promoted by this very medium.

I thought it suitably charitable to assume Stephens may have been too distracted to notice the fine detail, rather than being deliberately obfuscatory – his ‘bristling’ at the time of writing and all that.

The response via the ABC Religion account was as follows.

‘That’s dealt with in the article. Neither here nor there. No impact on the findings. And not an excuse for the crimes.^[7]’

I’ll take that as a negative on any confirmation of understanding, and no, it wasn’t ‘dealt with’ unless you call dismissing out of hand dealing with something. More followed…

‘Mmmmm… no, not quite. Maybe you should read the study. Regardless, there’s no evaluation being made here.^[8]’

You may recall the charge being made against the Australian media coverage, was self-righteous cynicism and wilful ignorance – the standard to judge by being ‘a lack of serious engagement’ with ‘substance and implications’. I won’t be giving prizes for people working out who, according to this standard, I think is being a cynical, self-righteous ignoramus.

Now you’ll remember I took issue with some of the report’s vague suggestions about potential causes of the crisis of child abuse in the Catholic Church, particularly those pertaining to historical moral upheaval.

‘For the Causes and Context study, the social indicators found to be most relevant to the modeling of the change in incidence of sexual abuse are divorce, use of illegal drugs, and crime. Sexual abuse of a minor by a Catholic priest is an individual deviant act—an act by a priest that serves individual purposes and that is completely at odds or opposed to the principles of the institution. Divorce is an act also made for personal reasons that negates the institution of marriage. Illegal drug use and criminal acts violate social and legal norms of conduct, presumably at the will of the offender.^[1]’

Forget technical language like ‘the institution of marriage’, it’s how the report substantiates a purported relationship with clerical child abuse that has me concerned. Or rather, how it doesn’t.

Citing Greenberg’s Time Series Analysis of Crime Rates^[9], the Causes and Context reports observes the statistical connection between divorce rates and two indicators of crime.

‘If the data for the annual divorce rate are compared to data for the annual rate of homicide and robbery, the time-series lines move in tandem. From stable levels in 1965, the rates increase sharply to a peak at or soon after 1980 and then begin to fall. This pattern is indicative of the period effects that can be seen in the Nature and Scope data on the incidence of sexual abuse by priests.^[1]’ – Emphasis added.

But just how ‘indicative’ is this ‘pattern’?

Greenberg, after using a Johanssen test to demonstrate cointegration, suggests a relationship between divorce, homicide and robbery.

‘It does not seem likely that this relationship is a direct, causal one or that people who are divorcing have exceptionally high divorce rates. More likely, divorce is an indicator of a strain in a fundamental social institution – the nuclear family. It is this strain that leads some individuals to kill, whether or not they themselves divorce. The divorce rate in the United States rose dramatically between 1960 and 1980, a time when gender relations and ideologies were undergoing major transformations, putting great strain on many families and leading some of them to divorce.^[9]’ – Emphasis added.

Gender relations and nuclear families – how exactly do celibate priests fit into the gender relations within nuclear families? I’m going to be brave and suggest an outlandish notion; that it’s a little difficult to experience the pressure of having a partner, two-point-four kids and a family dog during a period of social upheaval, when you’re a virginal bachelor.

Conveniently, the specifics of the relationship suggested by Greenberg remain unmentioned by the Causes and Context report.

But this isn’t the crux of my objection. It gets worse…

It’s been a common refrain in the criticism of the report in question, that old truth that correlation doesn’t equal causation. This is an important realisation, but I think it ultimately insufficient in appreciating what I take to be a fuller extent of the problem; establishing correlation is still useful, but in this case, questionable.

Unlike Greenberg’s Time Series Analysis of Crime Rates^[9], Causes and Context^[1] doesn’t even attempt to demonstrate a statistical relationship between the purported causes and effect. There isn’t a whiff of a regression analysis of any kind on the matter.

Whatever you may think of the relationship proposed by Greenberg, at least the data are associated statistically. Perhaps Greenberg’s suggested cause is true, or perhaps divorce, murder and robbery are together all driven by another unrecognised factor, perhaps something such as lead contamination. At least it’s demonstrated that the studied indicators move together (i.e. that they’re cointegrated).

It’s important to realise of Greenberg’s cited work, that owing to practical limitations, analysis of only two of six ‘index offenses’ are undertaken; robbery and homicide, from a list of offenses also including rape, assault, burglary and automotive offenses. Greenberg explicitly stops short of affirming the cointegration of these other indices (or by implication, any index other than the two analysed) with divorce rates.

Either as a sub-set of rape, or on its own, further analysis would be required to demonstrate a statistical relationship between child sexual abuse by priests, and factors such as divorce. Yet if the report team did undertake such analysis, the Causes and Context report singularly fails to document it. (Are the data even amenable to this?)

And no, simple visual inspection of a graph and pointing out that two or more things peak in the same decade does not demonstrate a correlation, not statistically at least. If correlations can be plucked out of the air so casually, we may as well throw out Pearson’s r and start scouring charts to find ‘correlations’ with fluctuations in bird migrations, trends in hairstyles, or the motions of the planets.

(Although the latter option, astrology, it has to be said, may at least attract Templeton funding, so there’s that).

Returning to the coal face at the ABC, how did Scott Stephens report the matter?

‘…those ordained before 1960 tended not to commit abuse until the 1960s and 70s, while those ordained in the 1960s and 70s tended to commit abuse very shortly thereafter. This would suggest that the foetid cultural soil of the 60s and 70s proved uncommonly conducive to the commission of sexual abuse.^[2]’

The jeremiad continues! Foetid cultural soil!

No, this does not suggest in any serious manner that a cause of child sexual abuse by priests is the culture of the time (or at least the parts thereof found unpalatable or otherwise convenient to blame). Not unless your definition of ‘serious’ includes wish-thinking.

I’m going to hazard a guess myself (because guessing is what’s going on) – that as a Coakley fan, pesky things like correlation coefficients are far too reductionist, too scientismist, for a sophisticated theologian like Stephens.

He goes on.

‘This line of reasoning has been characterised as the “blame Woodstock explanation,” designed to give the Catholic Church some alibi for its crimes. It does no such thing. Indeed, there can be no more damning indictment than that the Church had so imbibed the proclivities of the age that it reproduced them in its own life.^[2]’

What incredible narcissism! The great crime of the church is that when they drifted amongst the unwashed masses, their unsoiled, virginal feet touched the ground! The message to priests is that they must float just above the rest us; low enough to remain truly penitent, high enough not to get our muck on them.

I get the distinct impression Stephens believes the church occupies moral high-ground by default, a secular world tasked with proving otherwise – in purely ecumenical terms of course. What a merry-go-round we’d all have if we allowed ourselves to be set this task!

Critical thinkers will do well to recognise and avoid such pitfalls.

Whatever Stephens base-assumptions about the church morality may be, if the grandeur doesn’t give it all away, the credulity in uncritically wolfing down the ‘blame Woodstock explanation’ exposes Stephens’ fabulation.

‘…to discount what I have called “the foetid cultural soil of the 60s and 70s” as a factor out of hand, quite frankly, suggests an almost delusional belief in the health and progress of Western culture.^[2]’

 

What Stephens doesn’t seem to understand, is that the study doesn’t demonstrate a statistical correlation, much less the causal nature of a ‘foetid cultural soil’. (And no amount of Stephens’ penchant for citing sexual liberals justifying incest or pedophilia, such as in this case, Shulamith Firestone or Daniel Cohn-Bendit, will change this.)

So how to cap all this off?

‘The pope’s determination to purge the Church of what he has repeatedly called the “filth” of abuse and concealment, his pastoral care of so many of the victims of abuse, and his insistence on the Church’s “deep need to re-learn penance, to accept purification, to learn on one hand forgiveness but also the need for justice,” distinguishes him not merely as the person who has done more than any other to eradicate sexual abuse in the Catholic Church.

Benedict XVI is also the man who can best bring this desperately evil chapter in the Church’s life to a close.^[2]‘

Stephens started his article lamenting the poor standards of the media in its appraisal of the Causes and Context study, for not realising what he’s uncritically adduced from the report, and this he calls ‘self-righteous cynicism and wilful ignorance’. That he concludes not by reaffirming this criticism with the facts he’s asserted, instead diverting with his final paragraphs to argue the standard pabulum of apologia for the current Pope, is revealing.

By the end of this long line of sophistry he’s abandoned his original line of argument, no longer even bothering to pretend.

In undertaking this jeremiad, Stephens has demonstrated his own inability or unwillingness for ‘serious engagement’ with ‘substance and implications’. That the topic of discussion is so crucial, so incredibly serious, makes Stephens’ overriding ideological preoccupations all the more disgraceful.

Notes and References

[1] John Jay College Research Team (2011) ‘Causes and Context of Sexual Abuse of Minors By Catholic Priests In The United States, 1950-2010’, John Jay College of Criminal Justice, <PDF file: http://usccb.org/mr/causes-and-context/causes-and-context-of-sexual-abuse-minors-by-catholic-priests-in-the-united-states-1950-2010.pdf>.

[2] Scott Stephens (2011) ‘Catholic sexual abuse study greeted with incurious contempt‘, Australian Broadcasting Corporation.

[3] Miranda Celeste Hale (2011) ‘A worthless and dangerous report‘, Miranda Celeste.

[4] John Jay College Research Team (2004) ‘The Nature and Scope of the Problem of Sexual Abuse of Minors by Catholic Priests and Deacons in the United States‘, John Jay College of Criminal Justice.

[5] John Jay College Research Team (2006) ‘Supplementary Report: The Nature and Scope of the Problem of Sexual Abuse of Minors by Catholic Priests and Deacons in the United States 1950-2002‘, John Jay College of Criminal Justice, <PDF file: http://www.usccb.org/ocyp/JohnJayReport.pdf>.

[6] American Psychiatric Association (2000) ‘Diagnostic and Statistical Manual of Mental Disorders’, Fourth Edition.

[7] ABC Religion (2010) Twitter status, Australian Broadcasting Corporation.

[8] ABC Religion (2010) Twitter status, Australian Broadcasting Corporation.

[9] David Greenberg (2001) ‘Time Series Analysis of Crime Rates’, Journal of Quantitative Criminology, Vol. 17, No. 4.