This is a test (but not a trick)

I'm interested in understanding why MR has such a high-quality comments section.  I'd like you to consider this passage, from today's Guardian (not today's Onion), and try to write high-quality comments on it.

The statement, read out by Archbishop Silvano Tomasi, the Vatican's
permanent observer to the UN, defended its record by claiming that
"available research" showed that only 1.5%-5% of Catholic clergy were
involved in child sex abuse.

Let's see how you do.  If you can indeed produce high-quality comments, it means you're better than the other blog commentators.  If you can't, maybe it means that Alex and I are in some way better with regard to what we post and how we present it.  In that case, once our splendid framing is off-scene, you revert to your usual, rotten selves.  I want you to end up with most of the credit.


What is there to say? We could talk about the banality of evil and historical involvement of the Roman Catholic Church in it, make snarky comments comparing those numbers to others (Branch Davidians? Polish film directors?), or legalistic arguments as to whether the Catholic church in the US merits RICO prosecution and disbanding as a criminal organization (yes, probably, but it won't happen), but in the end the sheer awfulness of the status quo defeats us.

1.5% to 5%??? Considering how many priests there are, that is a really large number of sex abusers. I can't think of any quality control regimen that would say it's satisfactory to have a defect rate of 1.5-5%.

The pressure is getting to me already.
If this were a normal post, I would say this is an example of how the media, with help from interested parties such as anti-papists and trial attorneys, creates a distorted view of the world using statistics. The percentage is the same for other groups, as Dave says, but not just cult members. It's the same for protestant youth ministries, for example. Motivations to single out the Catholic Church are many (class bias, feminism, etc.) but the reality is child molesters are like any other criminals. They rob banks because that's where the money is.
How's that?

Now the meta level. Did you pick this at random? Surely not. In which case it is just another post of yours instead of a typical guardian post.

Do we get feedback on how we did?

I suspect that it would be more apropos to compare the rate of abuse between priests and the educated middle-class.

While I honestly don't know if abuse rates differ by education and social class, if they do, then priests would be better compared with their peers. (And I have no idea if they would fair better or worse.)

I will admit to also being curious as to how the Vatican came up with this 1.5-5% figure.

Also, according to "46,000 Catholic priests in the United States and 300 cases filed to date, which is under .7% of the priesthood". If the number is 1.5%, and the probability is no conditional upon country, we've only got half of them. If it's 5%, then we've got 2,000 more abusing priests out there to arrest.


Maybe the Catholic church should be recruiting Six Sigma black belts for their seminaries.

Okay, even MR comments sections have to have at least one idiot.

"Would it not be a better experiment to post this the quote and not let us know we are being tested?"

I agree with Andy.

The problem is that Tomasi is pushing a red herring here. The problem was never the percentage of sex offenders among priests (which is indeed comparable with other groups) but the incredible lengths that the church hierarchy went to cover up their abuses and protect the offenders thereby significantly increasing the damage done.

I also agree with Andy and Dr. Yogi.

"Would it not be a better experiment to post this the quote and not let us know we are being tested?"

Now we are biased to behave, or at least biased in some direction, depending on the person. Where is Robin Hanson?


Agree on wanting some more specifics from the Vatican on this "available research" it cites.

How do they find out who's "involved in child sex abuse"? Self-reported? Survey? (and if so, what kind of sample?) Consider only instances with filed police reports?

I believe you just biased the test by telling us to give quality comments. MR readers will now put forth unusual effort to provide high quality comments, while the readers of the Guardian did not know their posts would receive such scrutiny.

I was the child they used for this study. I put my foot down after 15 priests. Thus the wide variance. It was double-blind study, so I couldn't tell if there repeat offenders...due to the blindfold.

The "research" question appears to be is there a causal relationship between Catholicism and sexual abuse by its clergy. 1) Does the Catholic Church attract child abusers because it is the most populous religion and therefore provides opportunity to intimately interact with children--a selection bias? 2) Does the Catholic tradition of celibacy for its priests attract homosexual males who are also child abusers--another selection bias? 3) Or does the Catholic tradition of celibacy for its priests cause these men to abuse children?

I don't immediately see any way to identify a causal relationship between Catholicism and sexual abuse by its clergy. The reason is that even if the Catholic Church had a higher rate of child molestation it would be virtually impossible to deal with point #2. If there was an equal amount of abuse of females then that might be a different story....

As 'counter-attacks' go, this wasn't very effective. Claiming that 'only' 1.5-5% of an organisation are involved in appalling crime? This raises the question of whether such an organisation should be allowed to engage with schools & youth projects, let alone command moral authority. I hope the Archbishop is misled by the research & that the true numbers are lower.
And I suspect the nuance between paedophilia and ephebophilia are lost on the victims.

"available research"

Yeah. Available research shows that I'm the most awesome and attractive person in the world. I choose to believe it, and have no incentive to look into the mater.

Case closed.

I think you are just better with what you post and how you present it. The only thing I find interesting about the passage is what makes the "research" and its "availability" scare-quote worthy.

the joke of course is he almost never reads the comments anyway.
said as much in an interview somewhere

Since the question of "available research" has been raised, I would note that the John Jay Report estimated that number of priests accused of sexual abuse in the U.S. was around 4%.

These comments seem even better than normal. I think people are more responsive to your "let's see how good of commenters you are" challenge than they are to the actual content of the post.

Perhaps for a real test of your commenters you should post something very short, perhaps just a link with a brief sentence, and have some other blogs post the exact same thing.

1) The obvious: if the Catholic numbers are correct, the number is way, way too high.
2) If their Protestant vs. Catholic numbers are correct, the media may indeed be unfairly framing this issue as particular to the Catholic church, when in reality it is more widespread within religion. However, existing Catholic statements are highly political so the real story is hard to discern.
3) The Church should consider the effect that a vow of celibacy has on the applicant pool for the priesthood. In particular, the vow requirement would appear to disproportionately select people who either (a) have low libido [obviously fine] or (b) do not wish to express or explore their sexuality [could be problematic].
4) The Church should also consider the effect of a vow of celibacy during priesthood. Are priests beset by problems of sexual desperation?
5) If it is not already, comprehensive education on sexual abuse should be a mandatory part of indoctrinating children into the church.

let's assume the vatican is a rating agency and only 1.5-5% of rated securities turn out to be poorly rated. the lid blows off the 1.5%-5% and everyone mistrusts all rated bonds.

how did that work out?

in the vatican's defense, it would be hard to find every molester in an environment that demands no marriage (it is the vatican's demand though). we saw this with the rating agencies too. the vatican should be a major funder of brain scan lie detectors ala DARPA

The church tends to attract the dysfunctional (among whom include a small number of pedophiles) because it offers the promise of redemption, protection and forgiveness, plus the motivation to change. Unfortunately, human nature being what it is, the church environment also offers the ability to hide and the opportunity to continue past behaviors.

Does the "available research" say anything about the compared rates of child sex abuse in the churches of freshwater and seawater economics?

Regardless of whether the percentage of sex abusers is higher in the Catholic clergy than in other populations, the Catholic Church has a record of allowing abuse to keep happening by moving offenders from post to post and parish to parish. Even if only 1.5-5% of priests were "involved in child sex abuse," the rate of child abuse inflicted upon Catholics (and other individuals who encountered these priests) is bound to be much higher.

Ryan Tetrick, sure there is some hidden purpose ;)

Let us just pretend we are as intelligent as the top of the class MR commenters (or is it commentators?)

Sorry, but I have not found the many comments useful, and many comments were just veiled excuses to indulge in anti-Catholic venom.

For the scientifically-minded, use reliable and objective data on abuse by clerics in other Christian denominations and other religions; how do the statistics for Catholic clergy compare? Let's have a special sample of secular-minded elementary public school teachers for extra comparison.

My hunch is that the issue with the Catholic Church has not been a higher percentage of clerics who are abusive, but a lesser willingness to report the offenders to the police. We can check that. Let's have more empiricism and less attitude.

Do we need to consider sexual behavior with adults as possibly abusive? Or is the issue just children?

Do we treat 15-year-olds as children or adults, that is, should sexual activity between a cleric and a 15-year-old be considered adult consensual or child abuse?

The data I have seen uggest that the child abuse by Catholic male clerics is overwhelmingly homosexual. Would this fact build a case for much stricter screening by Catholic authorities to keep homosexuals out of the priesthood? Would those who say that are horrified by the priest scandals support such screening?

Just wondering: how many of the indignant leftists blasting the Catholic Church are also calling for the LA D.A. to drop extradiction on Roman Polanski?

This public confession from Archbishop Tomasi can be no surprise. For centuries, European popular cultures have been rich with jokes about clerical attraction to children. In English, a rhyme comes to mind attributing the phrse "jelly on springs" to a clergyman commenting on a choir boy's hind parts. In Spanish, an old joke about a choir boy's startlingly large contribution to a collection for the parish priest's new cloak turns on a a pun between the words for cloak and for castration. And so on.

As the Spanish story illustrates, the faithful used to care for their erring priests; even to collecting to clothe them more adequately. The Vatican's problem today is that the laity have become more moral. They now insist on following the Church's teaching that it were better for a man to suffer torment than to harm a child. The clerical heirarchy continues in the Church's tradition from more callous and permissive centuries.

Cowan's confession of surprise at the reputation of Marginal Revolution for thoughtful comment is equally no surprise to the rest of us. It illustrates the loss in appreciation of style that is inherent in his habit of at most skimming over what we write; at his incredible reading speed. However, he seems to have equally missd two relevant facts: his posts are too varied and interesting to attract the rut-bound trolls of the internet; and Tabarrok's are too well considered for them to get their minds around.

One could build an argument that goes something like this: 1. the catholic church attracts an unusually high share of homosexuals (example: ); 2. since homosexuality is considered a sin by the church, a culture of secrecy and guilt is fostered; 3. in this environment, the marginal cost to committing additional sinful sexual acts is lower than in the general population. The data cited by the article fails to be persuasive.

So, hoping that the actual statistic isn't as high in the normal population, does this mean pedophiles are attracted to religion? Is the statistic as high with teachers? Has the historic stigma of male teachers prevented such a scandal with teachers? Are we about to have a deluge of female sex abuse scandals as sex with female teachers becomes less excusable?

More generally, are we going to have to reevaluate sexual norms in light of a huge population of sex offenders existing - assuming the statistic is played down (I can't imagine he exaggerated it).

Also, why would he make this comment? I don't think anyone thought the statistic was that high. Is the Catholic churches PR really that out of touch?

I wonder if there is any literature on how physical and sexual abuse cascades from the abuser, to the abused, to the children of the abused and others in his life. Obviously it won't be as clear as tracking the spread of a disease, but I'd like to know just how much impact these 6,000-20,000 abusive priests have had.

How many victims does the average child abuser have?

When I see numbers like this I can't help but to want to reverse engineer the source. My guess is the range just came from a sample average with 2 standard deviations in either direction. By my back of the envelope calculations I get 3.25% avg and 0.875% sd suggesting they took a sample of 400 members and found 13 offenders.

Was that sample size sufficient? Well, if they just wanted to show that the percentage of offenders is well under 10% then they accomplished that (assuming we are confident they took a truly random sample). However, I tend to think that if the upper bound of your confidence window is more than two times the lower bound (when estimating a percentage) then you've probably taken too small of a sample to do any type of useful modelling; unless your lower bound is much closer to 0 than it is here.

I thought the Church was supposed to be the Guardian of ethics and morals. Once peeled away like an Onion , it brings tears to the eyes.

I'm struck that so few responses offer a solution. This New Yorker article written by Malcolm Gladwell may suggest one.

To me it's self-evident that sexual abuse by priests is a small part of a much larger problem. The Catholic church can choose to be part of the solution though the one suggested by Gladwell likely would not be how they'd go about it.

Citations on the table please... don't tell us waht the "available research" tells us. Show us the research. Is this research published? Was it made up? Give us sourcews and let us judge the quality of the "available research."

Paris Hilton is hot.

Several posters here are pointing out that Tyler set this up rather poorly if he wanted to use it as a test in the manner he outlined.

Since it seems so obvious that it is a poor way to go about it, "quiz factor" makes me wonder if perhaps Tyler is engaged in a sneakier experiment of some sort.

Thus perhaps the proper question we should all be trying to answer is, "What is the real purpose of Tyler's post?".

Anyway, regarding the nominal issue at hand, I think that while there is some small amount of outrage that some amount of Catholic priests engaged in pedophilia, I think that the greater outrage that people feel is that instead of (quietly, presumably) kicking these priests out of The Church, they simply took them from one parish and put them in another, and rinse and repeat.

Practicing pedophiles are perhaps the most reviled creatures amongst humanity. However even worse than that, to a higher power (so to speak), are those authorities who chose to turn a blind eye and allow it to continue again and again.

As far as I am concerned, this is one of the most pointless postings I have ever seen here. What an
utter waste of everybody's time.

Quality enough for you guys?

Based on a table/formula I found (, a batch of randomized, thorough investigations (Inquisitions?) of 2,200 clergy would generate a margin of error of +/- 1.75% at 90% confidence level (5,350 at 95%). Such an effort would be rather expensive.

I'd like to know the incidence of deviant behavior in the general population, or within various demographics that are present in the priesthood (age, ethnicity, location, etc).

As others have pointed out, the figures cited probably refer to those involved in covering up or negligently failing to take action, rather than the actual abusers. However, the phenomenon of "closing ranks" in the face of bad, corrupt or even merely incompetent actors within one's own group is pretty much universal (eg, teachers' unions, police unions, the medical profession, etc).

The Catholic church was probably hardly any worse than other institutions with staff in a position of authority over children (orphanages, boarding schools, juvenile detention facilities, etc). That is not an apologia for the church, by the way, but rather an indication of just how bad and pervasive abuse of children was, within living memory.

So why then was the church more or less singled out? Quite simply, as a monolithic hierarchical global organization it could be held liable for the actions of any of its "employees", and thus made an extremely tempting deep-pockets target for litigation. Furthermore, it was also an organization whose top leadership was guided by moral principles and therefore capable of feeling shame, and willing (even eager) to change and atone for its past bad behavior -- unlike, oh say, Wall Street banks -- which made it more likely to eventually cave in, quit stonewalling, and agree to very generous financial settlements. Lawyers predictably took the Willie Sutton approach in selecting which injustices from the past to drag into the light of day.

OK, there is always another look at things. By claiming that 1.5%-5.0% is small ("only"), the Vatican is covering one aspect of the story - "Gee, those priests are bad." The answer is "See, not soooo bad." But what remains unexamined is the process the Church put in place for dealing with predators. The fact that an institution which presumes to tell us how we may reproduce, how we may approach god, and what form worship must take is also in the business of sheltering criminals, exposing new communities to those criminals and making claims about the goodness of those criminals might be a bigger issue than whether 5% is a big number of a small one.

I don't know how many pedophile priests the Church identified - though there is evidence it didn't try hard to identify them - but we know that those who were identified tended to be pulled out of the community in which they had committed a crime, repackaged and sent back out to new communities which had no knowledge of their crimes. So while our attention is drawn to the "fact" of a low number (notice that many of the comments here are about whether it is a low number - success!), the institution drawing our attention has its own crime to answer for. The child-rapist-repackaging institution needs us to take for granted that its community-level operatives are statistically unlikely to rape our children, in order to maintain its own status.

One final point. We may well see priests as people who have chosen a certain employer, way of life and such - ordinary, analogous to being a lifer in the Army - but that isn't the story they tell. The Church maintains that priests are extraordinary. They claim to be the way to salvation. The Church still claims that its way of seeing moral issues is the one true way. If the priesthood has only the same share of child-rapists as the general male public, is that really reason to say "Oh, well then go about your business"?

Oh...hope that was civilized.

Those Who Know routinely assure me that almost all child abuse is "within the family", meaning that it's done by fathers to their children, or perhaps by stepfathers or live-in lovers to the children of their wives or mistresses. So the figures for the RC priesthood presumably need mainly to be compared with the minority of cases which are committed by non-family members; in which case, 5% seems pretty high.

The more interesting question is "what percentage of reported child abuse cases involve Catholic priests?"

Any thoughts on whether the view as to the priest's infallibility matters? Do religions with less infallible leaders lead to lower levels of child abuse? Also, the self selection hypothesis could be tested by looking at instances at boarding vs. non-boarding schools.

It is really too bad that Vatican is locking up a data set that could help us understand the economics of pedophiles (Mega Freakonomics?).

I would like to see a studying comparing sex offenses of Catholic priests to other professions (like other Christian ministers, school teachers, etc). I don't know whether 1.5-5% is high or low relative to other similar professions. I suspect that it's high.

Is that high or low relative to the general population?

If it's high relative to, say, other Christian ministers, how much do their vows have to do with it? Do they commit these crimes out of some extreme form of repression or do they self-select into this particular profession?

Some carefully designed field experiments and/or some good data could get to this bottom of this...Mr. Levitt, where are you?


The Catholic Church needs to take a six sigma class on quality control!

1. Since nobody went low-brow: Polanski says to the priest, "that's what the hot tub's for!"

2. People have asked about how much of the clergy sex abuse is a gay problem, but I don't think anyone here has focused on the age of the victims. People who rape 6 year-olds are probably very different from people who rape 16 year-olds. Which group are rapist-priests more likely to go after?

For everybody who suspected this was a trick despite what Tyler said: the answer is two posts below.

As a psychologist, I think I can safely say this is the most confounded behavioral experiment I've ever seen undertaken with any degree of seriousness.

This experiment typifies the problem of many behavioral economics experiments.

Agreed Andrew John. The comments here generally are fairly typical for an economics blog -- 10% insightful, 60% redundant/pointless and 30% dogmatic paultard astroturfing.

I'm still stuck, jaw gaping, at the "only..." portion of the quote.

The off-handed and dismissive tone that can easily be inferred from that statement is quite a slap in the face.

More significant than the population stastic you reported is this-- each child was 100% abused.

I, too, am curious about the meta-level. Why would you inform us of the intention of the post, rather than simply posting it and analyzing the results? Every freshman psych student knows that subjects modify their behavior when they believe they are being observed. Secondly, by specifically soliciting comments you have increased the number of responses, which could have one of two effects: by people trying harder to comment intelligently the average level of comment might be better than normal (unlikely, in my view), or the increased number of comments will lower the average quality (given that under normal circumstances a self-selected level of decent commenters choose to comment, whereas the self-selection in this case has been altered by your intervention).

Surely, you, Tyler, have thought of all these effects (and perhaps many more), yet you chose to let us know we are under observation. What is your game, professor Cowen, and why have I spent 5 minutes of my life trying to deduce your true motive?

IMO Tyler was trying to get us to talk civilly about this topic.

It is very hard for me to believe that 1.5% of the general male population are child molesters. So does the priesthood attract them, I could see that, or make them, I could also see that, or both.

Why do you say this blog has better quality of comments? I mean compare to which blogs? My theory is blog of certain subjects always gets only people interested in that subject and hence comments they will provide will be more appropriate to blog entry. Rents and YouTube video kind of entries get comments from all over and hence poor quality.

Regarding whether this should have been a blind test: there may already have been a blind test.

Here's a meta comment. The number of comments on this thread is significantly higher than on most other threads. That means the likelihood that the post quality is >= average is very, very low. And all the more impressive a statement about commenter quality if it stays high.

It's the psychological Heisenberg uncertainty principle at work. You can't measure something accurately once you ask people about it.

Moral concepts are best thought of as matters of absolute rather relative performance. To accept relative performance as the basis for moral judgements leads to conclusions that something is OK simply because everyone else accepts it. Some of the darkest moments of history are characterised by judgements along those lines.

Ostensibly the Catholic Church is an institution concerned with morality. It is jarring to see a morality-based institution furnish statistics in that fashion. It demonstrates its fundamentally relativist, defensive nature on this issue.

Ideally, our moral institutions would accept their moral challenges in absolute terms, without qualification, and be accompanied by pro-active intent rather than reactive, defensive responses.

Once again, a bad defense is worse than a skilled attack. If the best that an Archbishop can come up with is "only" 1.5%-5% there is something seriously wrong happening. I always figured Catholic child sex abuse was a big deal because of the hypocrisy, complicity, and weirdness of it all. But whoa. This makes me think it was a widespread and shockingly common phenomenon. Seriously, this completely changes the way I think about the issue in a way an outsider never could.

Comments for this post are closed