Monday, July 30, 2007

Religion and Crime

I've been reading the literature on the relationship between religious conviction and crime, as part of my thinking about the relationship between philosophical moral reflection and actual moral behavior. The literature is pretty weak. Much seems church-inspired and probably deserves about the same level of credence as drug-company funded research showing their blockbuster drugs are wonderful. Much of it is in weird journals.

I found a 2001 "meta-analysis" (Baier & Wright) of the literature that shows all the usual blindnesses of meta-analyses. Oh, you don't know what a meta-analysis is? As usually practiced, it's a way of doing math instead of thinking. First, you find all the published experiments pertinent to Hypothesis X (e.g., "religious people commit fewer crimes"). Then you combine the data using (depending on your taste) either simplistic or suspiciously fancy (and hidden-assumption-ridden) statistical tools. Finally -- voila! -- you announce the real size of the effect. So, for example, Baier and Wright find that the "median effect size" of religion on criminality is r = -.11!

What does this mean? Does being religious make you less likely to engage in criminal activity? Despite the a priori plausibility of that idea, I draw a negative conclusion.

First: A "median effect size" of religion on criminality of r = -.11 means that half the published studies found a correlation close to zero.

Second: And that's half the published studies. It's generally acknowledged in psychology that most studies that find no effect -- especially smaller studies -- languish in file drawers without ever getting published. Robert Rosenthal, the dean of meta-analysis, suggests assuming for every published study at least five unpublished studies averaging a null result.

Third: As Baier & Wright note (without sufficient suspicion), the studies finding large effects tend to be in the smaller studies and the studies co-ordinated through religious organizations. Hm!

Fourth: The studies are correlational, not causal. Even if there is some weak relationship between religiosity and lack of criminality, some common-cause explanation (e.g., a tendency toward social conformity) can't be ruled out. Interestingly, two recent studies that tried to get at the causal structure through temporal analyses didn't confirm the religion-prevents-criminality hypothesis. Heaton (2006) found no decrease in crime after the Easter holiday. And Eshuys & Smallbone (2006) found, to their surprise, that sex offenders who were religious in their youth had more and younger victims than those who were comparatively less religious.

Does this suggest that religion is morally inert? Well, another possibility is that religion has effects that go in both directions -- some people using it as a vehicle for love and good, others as a vehicle for hate and evil. (Much like secular ethics, now that I think of it!)


Anonymous said...

Saw this in the news today:

"Religious doctors not more likely to care for poor";_ylt=Aj.ProoAhqG4yq8SNMAiht0DW7oF

Anonymous said...


The old brilliant cumudgen, sociologist Phillp Rieff, shortly before his death, wrote a book entitled "Charisma: The Gift of Grace and How It Has Been Taken Away from Us." In that analysis he deals with the issue of whether religion (primarily Christianity) is morally inert or not.

He argues that the Christian faith tries to reorganize the passions in the following way: "doing something else (y) when something (x) is done to you. In the Christian symbolic, the Passion of Jesus opposes the more familiar passions when he gives up his life as a ransom for the many and thereby takes an order of action beyond fear and anger--to guilt.

Rieff sees this action by Christ as the passion to end all other passions--becaise, for Rieff, all freedom to be only passionate will end in the most terrible transgressions.

The institutionalzed Church, through the sacrements, also for Rieff, intensifes the consciousness of guilt at the same time that faith is rendered triumphant.

Rieff sees modern American culture as broken. What passes for authority in our country has lost its capacity to engage in its own discipline of obedience to a set of thou shalt nots. And with that type of refusal all we are left with is power and endless war.

Eric Schwitzgebel said...

Timely, Boram! Thanks for the link! Here it is, a little easier for others to use: Religious doctors not more likely to care for poor.

Jim, I'm not sure I understand the bit about the passions, but (despite appearances?) I'm not possessed of a very cynical, curmudgeonly disposition. I think there are very many positive things in contemporary American culture, both secular and religious, and not all is power and war. But the relation between what is good in us and philosophical reflection or religious avowal -- that I'm not sure of.

Unknown said...

Addressing what it is to have "religious conviction" seems likely to be problematic for any study. I'm sure a high proportion of people in the US affirm that they are religious/spiritual, even many of those sitting in prisons today, but I suspect few of them have very serious convictions. (OK those in prison have convictions, but not the appropriate sort!) I wouldn't expect there to be much variation between the religious believer/non-believer when the former is of the tepid, nominal, weak, sort. However, I'd would be surprised if the same held with those that were of serious disposition.

Anonymous said...


At least some of the data looks not merely at professed belief, but religious attendance. See, for instance, Paul, G. 2005. Cross-National Correlations of Quantifiable Societal Health with Popular Religiosity and Secularism in the Prosperous Democracies. J. Religion and Society. Paul claimed to find a negative relation between religiosity and morality

Eric Schwitzgebel said...

Matthew, you raise a good point. Researchers have looked at this in a number of ways, both in terms of self-report of religiosity, or degree of religiosity (since few people profess to be out-and-out atheists), and others looking a more objective measures -- as Neil mentions -- like church attendance and membership. It doesn't seem to make a big difference in the results.

Anonymous said...

This is an immensely interesting topic for me, but one that I find also immensely subtle and complicated. Any good statistics contrasting religious with non-religious would be very helpful – e.g. it would be nice to know if getting baptized made you 500 times less likely to commit adultery – but I only think this can get us skin deep. I agree with one commenter on this thread that our culture is a bit nuts for this kind of analysis, and mere ‘religious’ vs ‘non-religious’ might be too abstract to get at any real causality. For example, I was once affiliated with a communitarian group with a deep narrative, and yet one that was fairly sophisticated (they recently hired a Phd in philosophy from Notre Dame). And last I checked, their divorce rate, calculated over the course of 15 or 20 years, was 6 percent. For this community to perform in this way required religion in my opinion, but I think the situational factors responsible can be discussed in non-religious terms, in so far as they incorporate social mechanisms that can be manipulated without the explicit use of religion.

I wonder if something similar could be said about doing ethics vs. doing good.

Michael Metzler

Eric Schwitzgebel said...

Thanks for the comment, Michael. I agree it would be interesting to look for differences among different types of religious groups. Unfortunately, there just aren't the data out there to start drawing those sorts of inferences.

(Part of the problem is that I think one should never believe just one study.)

Anonymous said...


Your comment about not being sure of the relation between what is good in us and philosophic reflection or religious avowal is indeed a key issue.

What I have find most useful is careful reflection on a passion or impulse (such as personal hatred or anger). My own anger and hate seems to have at its foundation a great deal of fear. The acknowledgment(through careful refection) that much of my anger is fear and that such fear is often transitory, opens up, at least the possiblily, of not acting on such impulses, as seeing these impulses as not necessarily me.

As anger and fear become more and more not-me, love and joy seem to become more possible.

Reflection becomes an internal check on anger and fear and the more they are checked the more love becomes possible.

Such reflective experiences can quickly bring one to ones knees, and such kneeling seems to be a recognition of a power superior to one's own.

Jumping to the marco-level one could argue, as does Rieff, that cultures are organized around the emotions of love and hate. His interpretation of Christianity focuse on how it has tried to reorganize the passions by doing somethng else (y) when something (x)is done to you. For Rieff, without a belief in the sacred we are left with only politics, whose essence is to become the x who does onto others before Y is done to him.

If both Religion and philosophical reflection can bring us to our knees it is, perhaps, because both ultimately require obedience to an authority greater than ourselves.

Eric Schwitzgebel said...

All that has some plausibility, Jim. But now let's see the empirical data!

Reactor said...

If you want a big study with a sample of 50 million plus you can use the UK consensus and the jail figures, it clearly shows that those who don't believe in God are more likey to be in jail.

Analysis of the figures show the non-reigious are twice as likely to be in jail.

15.5% non-religious in the popuulation and I estimate. 33.9% non-religious in jail.

15.5% non-religious in the popuulation and I estimate. 33.9% non-religious in jail.

Eric Schwitzgebel said...

Thanks for the links, Reactor. Two methodological issues come to mind: One is whether the questions used the same wording. (They're from different years, so presumably not the same survey.) Another is what confounds there might be. For example, low-SES people or people of a certain ethnicity might be more likely to be in jail and also more likely to say they are not religious, even if *within* each group studied, the religious are just as likely to be in jail.

In general, confounds are extremely problematic in studies of the relationship between religiosity and crime, since religion cannot be randomly assigned. The better studies control for some confounds, but there are always many potential confounds that cannot be controlled for, such as (for example) having a conformist personality, or high (academic) intelligence, or....