In an interesting study titled "The Negative Association between Religiousness and Children's Altruism across the World," led by Dr. Jean Decety of University of Chicago, a team of researchers in psychology found that children in religious households tend to have less willingness to share what they have with others than children in non-religious households.

Religious people have reacted predictably, calling the study "unscientific" and "pure crap," these judgments coming from people who, as far as I can see, have not specified what their credentials are for evaluating papers published in Current Biology, a peer-reviewed scientific journal. For those not familiar with the publication of scientific research, "peer review" is the process used for judging whether a scientific research paper is worthy of publication in the journal to which it has been submitted: the paper is sent out to recognized experts in the field who review it, point out any flaws for correction, and who may reject the paper if it does not meet the standards of that field of scientific study. For the journal Current Biology, the Information for Authors describes this process in detail. The paper "The Negative Association..." passed through this process.

The lead author of "The Negative Association...", Dr. Jean Decety, is the Irving B. Harris Distinguished Service Professor of Psychology and Psychiatry at University of Chicago. For those unaware, it is a very high professional honor to have a named professorship at any university, let alone one of the most prestigious research institutions in the United States. That doesn't end the list of Dr. Decety's professional accomplishments: see his page here on the U of Chicago website.

So I don't pay much attention to people saying the paper is junk science written by hacks who don't know how to do research.

Here is a good non-technical summary of Dr. Decety's paper on the U of Chicago website. To quote the central findings of the paper from this summary: "Children from religious families were less likely to share with others than were children from non-religious families. A religious upbringing also was associated with more punitive tendencies in response to anti-social behavior."

The study included 1170 children, from six countries (Canada, China, Jordan, South Africa, Turkey and the United States). For each of these children, the religiousness of their home was assessed, by several different measures (specified in the paper), and the children were interviewed in their native languages by trained researchers, who administered tests of moral sensitivity and had them play "The Dictator Game," a measure of altruistic behavior that assesses a child's willingness to share possessions with another (anonymous, unseen) child. (I will say more about this "game" later.)

There were three groups, among the children, that were large enough to make it possible to reach statistically meaningful conclusions about those groups: Christian (280 of the 1170 children), Muslim (510 of the children), and Non-Religious (323). That accounts for 1113 of the children, so no meaningful conclusions can be made about, say, Hindu children alone, or Jewish children alone, in comparison with any of the other groups, because there weren't enough children of those religions to make it possible to reach any reliable conclusions on those religions by themselves. (That does, in fact, accurately reflect the fact that the great majority of people in the world are Christian, Muslim, or have no religion.)

The average scores of the children from non-religious homes on the "Dictator Game" showed a higher willingness to share possessions than the average scores of children from Christian homes or Muslim homes, or from "religious" homes in general. The difference was, to use some statistical jargon, "significant at the p<.001 level" -- translation: the finding, in the course of the research, of a difference between religious and non-religious children in their willingness to share possessions had less than one chance in a thousand of occuring just by chance. Most journals will accept a finding if the likelihood of it occuring by accident is less than 5%; in the case of the finding in Dr. Decety's paper, the chance that the distinction between religious and non-religious children's willingness to share is an accident is less than 1/10 of a percent. It seems non-religious children really are willing to share their possessions more than religious ones are.

This is not what Christians or Muslims would claim to be the case.

If you are reading this, and are making judgments of your own on the research, positive or negative, those judgments are premature. So far, if you have followed the link above to the university website summary (and certainly if you haven't followed the link), you haven't seen enough. You haven't seen the research; you've only seen what a writer (Susie Allen) said about the research. (Or what I have said about it.) No judgment can be valid until you read the research paper itself, and it can be found here. (You need to purchase the right to view the article in order to read it. I read it while it was still available for free.)

The paper (on page 5) only very briefly describes "The Dictator Game," probably because the authors assumed that people reading the paper are familiar with it, but it does give a reference for the "game" that explains it in greater detail: "Children's altruistic behavior in the dictator game", a paper from 2007 in the journal Evolution & Human Behavior, introduces the "Dictator Game" and describes it completely. It is intended as a measure of altruism in children, and it consists of showing the child thirty very pretty stickers, asking the child to choose ten that he/she likes best, telling the child he/she can keep them, and then saying: "I only have time to give stickers to some of the children in your class, but not to all them. If you want to, you can give some of your stickers to a child in this class whom I do not give stickers to. You do not have to give any of your stickers away, but if you want to, you could give some to a child who didn't get any. I do not know which child will get them, and you will not know. Another lady will decide who gets them later."

(In the "Children's altruistic..." paper, read the section "2.2: Procedure." It says a lot more than I just did. Care is taken to ensure that the child understands what the deal is, that the child understands he/she is under no obligation to share, and that the child knows that the interviewer sitting in front of him/her will not even know whether the child decided to share any stickers or not. The procedure includes safeguards against, say, a child setting aside stickers to share because he/she believed the interviewer expected him/her to do so, or would be mad if he/she didn't.)

This is important: in the "religious altruism" study by Dr. Decety, he and his group did not invent this measure of altruism for the purpose of their study. It was invented years earlier, by an entirely different group of researchers for an entirely unrelated study. Dr. Decety's group used it because it has, by now, become an accepted way of measuring altruism in children. I am saying that to undercut one of the amateur criticisms I have read.

Let me undercut another criticism, while I'm at it. I read one reaction that said that Decety had not taken different national traditions or rich-kid vs. poor-kid into account. That is, suppose the atheist families were mostly in China. In that case, the study would be confusing atheist kids' willingness to share with Chinese kids' willingness to share, unable to tell them apart. Similarly, perhaps more of the Christian kids were from poor families, and were less willing to share what they had because they were poor, not because they were Christian.

That critic of the research simply hadn't read Dr. Decety's paper, or perhaps read it but didn't understand it. Let me quote from page 3: "Results from a linear regression with number of stickers shared as the dependent variable and age, country of origin, socieconomic status, and overall religiousness of the household, suggest that age and religiousness were significant predictors of sharing." I don't blame anyone for needing to have that seeming gibberish translated into understandable English, because the number of people who have studied graduate-level statistics at a university is small, but, as it happens, I am such a person, and I know what that sentence meant: that the mathematical analysis done by Dr. Decety's team separated the effects of age, wealth, nationality, and religiousness on a child's willingness to share, and found that children of the SAME AGE, SAME COUNTRY, AND SAME SOCIOECONOMIC STATUS were more willing to share if they came from a non-religious household than if they came from a religious one. The critic who said that the study was overlooking factors such as national culture or rich-vs-poor was simply wrong: the authors looked at all those factors and took them into account.

The conclusion flies in the face of the near-universal belief among Christians and Muslims that atheists are immoral people, not to be trusted, that atheists are not motivated to exhibit good behavior because they have no fear of God punishing them for behaving badly. It is not surprising that so many Christians and Muslims instantly reject the research performed in the study and the qualifications of the research team to perform it, without actually reading the paper or checking on the background of any of the authors.

I have a theory of my own to explain the results of the study. I'll state my observation in a picturesque way, and then explain it afterward. I think that religious people (particularly Christians and Muslims, who believe everything is controlled by a single supreme being, the God of Abraham) see their most important connection to the world as vertical, while non-religious people see their most important connection to the world as horizontal. And that is what causes the results of Dr. Decety's study.

(What in the hell do I mean by that? Bear with me, please.)

Christians and Muslims see their relationship with their God as the most important fact, almost the only important fact, of their existence. That relationship is the key to what they want above all: ensuring that their life goes on forever. They have relationships with other people around them but, aside from Christians wanting all those others to be Christians and Muslims wanting them to be Muslims, those surrounding people are far less important than God. (That is the vertical relationship I was talking about: between self and God.)

Non-religious people have no relationship with the Man Upstairs. Their most important relationships are horizontal, between themselves and all other people in the world around them. Non-religious people teach their children to care about other people, because there is nothing more important outside themselves (such as God) for them to care about.

I believe that, as a consequence of that, the motivation for a non-religious child to share his/her possessions with others is greater than for a religious child, a child taught by religious parents: It's not that religious parents don't teach anything about sharing, on the grounds that sharing is something God would like to see you do, but for religious parents is just one of many, many things to teach the child, mostly having to do with God. Non-religious parents don't have all that teaching-about-God to worry about, which increases the focus on the relationships with other people, and the importance of sharing with those other people.

I'm not going to insist on that as an explanation. It's just a thought.