An interesting report published in Current Biology 25 (Decety et al., The Negative Association between Religiousness and Children’s Altruism across the World http://www.cell.com/current-biology/pdf/S0960-9822%2815%2901167-7.pdf) suggests that there is a negative association between religiousness (of household) and children’s altruism across the world. This contrasts with a higher parental perception (in religious households) that their offspring are empathic and sensitive to justice, and a more judgmental and punitive attitude to interpersonal harm amongst the children themselves. The Guardian writer highlighted altruism (good) and judgmental (bad) in the summary line, suggesting (as do the researchers) that the long-standing belief that religion contributes to moral development is misplaced. The implication might be taken to be: remove religion from the human educational and socialization process, and there will be a world-wide increase in altruism and decrease in judgmentalism, at least in primary age children.
For those that have a visceral dislike of religion and all its works, and would like to see it removed from the educational system as soon as possible, this is powerful grist to the mill. At a deeper level it provides a welcome challenge to the unthought-through bleatings of some religious people (many of whom should know better) that all non-believers are ipso facto selfish, individualistic materialists who don’t know right from wrong. Not only are non-believers not selfish after all. They turn out to be statistically even more not-selfish than their religious detractors.
The first thing to say is that as far as I can judge (my grasp of stats is a bit hazy, though I am working on it) the figures are, in the words of the researchers, robust. The interest is then on looking a little more closely at the story that they might be trying to tell. Please check my argument against the article carefully, because I might miss some of the detail, in which case, of course, my comments will not apply.
I will declare my special interest at this stage (this will be obvious from other things on this site). (1) I find the in-principle claims sometimes made by religious people who assume the moral depravity of non-believers irritating, ignorant and deeply unhelpful for any fruitful conversation around vital issues of communal living in a complex society. (2) I believe the religious ethical tradition that I am at home in, which might be (accurately but cumbersomely) described as critical-modern Judaeo-Christian humanism with classical roots, continues to have something to offer in the education and socialization of young human beings.
Let’s look at the question of altruism. How was this tested? Well a version of the Dictator game was used – a not-quite-game theory decision test whose results have been used to challenge dominant assumptions about economic rationality. I have just experimented with a version online, and in my efforts to be nice to the peasants have had to survive two assassination attempts by the army in my first four months. I might come back to that. This is of course a remote scenario for the normal five – twelve year-old. It’s pretty remote for a normal Jesuit supply-teacher as it happens. So, in the language of the researchers they employed an ‘ecologically valid’ test:
“In this task, children were shown a set of 30 stickers and were told to choose their ten favorite. They were then told ‘‘these stickers are yours to keep.’’ Children were instructed that the experimenter did not have the time to play this game with all of the children in their school, so not everyone would be able to receive stickers.”
“Of additional note is that the sharing of resources was with an anonymous child beneficiary from the same school and similar ethnic group” [avoiding ‘in-group’, ‘out-group’ biases].
And here are the results:
“Results from a linear regression with number of stickers shared as the dependent variable and age (1-year bins), country of origin, socioeconomic status (SES), and religious identification of the household (dummy coded) suggest that age (bstandardized = 0.39, p < 0.001), SES (bstandardized = 0.16, p < 0.001), country (bstandardized = 0.1, p < 0.01), and religious identification (bstandardized = ␣.132, p < 0.001) are significant predictors of sharing, (model r2adjusted = 0.184). Paired comparisons (corrected for family-wise error) showed that Christian children (Msharing = 3.33, SD = 2.46) did not differ in their altruism from Muslims (Msharing = 3.20, SD = 2.24); however, both were significantly less altruistic than non-religious children (Msharing = 4.09, SD = 2.52, both p < 0.001”
It is not completely clear to me how the task worked, but it looks as though, having been allowed to choose their ten stickers, the children were given the option of sharing some of their selection with an (unknown) person who otherwise would not receive any. The average religious child shared 3 1/3 stickers and the average non-religious child, just over 4. The importance here is that this was not a hypothetical experiment (as in my dictator game – I might well alter my behaviour in real life, after the first assassination attempt), but a real choice of the children concerned, and thus a real cost. The numbers: 1170 children across six countries (US, Canada, China, Turkey, Jordan, South Africa (Cape Town). Roughly 310 would count as non-religious.
It is apparently suggested statistically in other studies that the capacity for altruism increases between the age of five and twelve and this was confirmed in the correlation coefficient of 0.39 above. Other significant variables were country of origin (0.1) and social economic status (1.6). When these were accounted for, the religious households showed a negative (-1.32) correlation, getting relatively worse with age, over against non-religious ones, midway between country of origin and socio-economic status.
There are follow-up questions for this experiment. How does the experimental behaviour of the children in the sample compare with their behaviour in the playground? What would happen if it was about sharing food, rather than ‘property’ or ‘gifts’? What form will their teenage rebellion take, and will it leave them more or less committed to the rules of their parents? More or less resentful of a demanding world? More or less autonomous in their treatment of rules? How will their education (religious or non-religious) develop their compassion and their capacity for deep and subtle discernment of difficult moral questions? What sort of adults will they turn into in response to the kindnesses and cruelties of daily life and their fellow human beings? Longitudinal studies would be interesting here. Nevertheless there is, prima facie, an interesting case to answer: why might the religious-household children, at this stage in their lives, other things being equal, give up fewer of their stickers? Or (a slightly different question) why might the non-religious-household children give up more?
The first thing to say is that the answer is very unlikely to do with a basic belief in the existence of God or the presence or absence of religious experience. Such belief or experience is in itself is behaviour-neutral, though (in the case of a conversion experience) it can lead to behaviour change. However, it may very plausibly be to do with the things that are believed and communicated by parents about God and what they believe is at the heart of their religious tradition. Note here that religious households can be very different in what exactly they communicate. This should not be surprising since most religious textual traditions alone tell multiple and conflicting stories. Is your God primarily the figure who will obliterate you if you step out of line or the figure who loves you with an everlasting love?
So let us take as a working assumption that all households generally will communicate in some form a big story about the universe and other human beings in relationship to rules, behaviour, reward and punishment, and this will give the conceptual framework in which children develop their thinking about behaviour. But alongside this, the specific behaviours modelled by elders and betters, and the daily messages (do X, don’t do Y, be like Z) delivered with (or without) consistency and fairness will be key factors.
The second thing to notice is that the prima facie argument supported by the statistic is not that religious children are not altruistic, it is that they have a tendency to be less altruistic than their compeers. This may mean either that their natural capacity for altruism tends to be hindered by growing up in their environment, or that their fellows grow up in an environment that enhances their altruism. These are slightly different things.
Let’s start with rules, because other things in the report suggest that these may be a key factor. A second set of statistics looks at making judgments and prescribing punishment. It shows a significantly higher correlation between a negative judgment about harming other people (“that’s mean”) and religious background. It also suggests (at least in some religious communities and in some cultural contexts) that religious children tend to prescribe harsher punishment for misdemeanours.
A slight digression here. Both the report itself (in its charts) and the Guardian in their discussion conflate these two rather separate activities of identifying a wrongdoing and prescribing a penalty in order to apply the ‘judgmental’ label to religious children. But it is significant, for instance, that – according to the report – the group of christian religious children was not significantly different in awarding penalties from the non-religious group. This suggests that ‘religious’ is indeed a critical variable for the former, but that other factors (including variations of interpretation within religions and cultural parenting practices) might play a more significant role in the latter.
The statistic that religious children are quicker to identify and condemn behaviour that hurts other people as ‘mean’ might plausibly suggest that the religious group are living in an environment in which their attention is regularly drawn to particular behaviour types – possibly with some discussion of the why’s and wherefore’s (hence the ‘that’s mean’ judgment). The fact that children from some religious households in some countries make harsher judgments is (I suggest) most likely to reflect the experience of correction at home, reinforced by a particular narrative of reward and punishment. That probably needs further investigation.
To report the first of these on its own as an index of how ‘judgmental’ a child would be a little misleading, since that term is usually pejorative. It has connotations of a lack of mercy and understanding towards someone who needs and deserves it. It might well be true that children from a certain type of religious household might turn out statistically to be less forgiving than others, but that would require a separate and more subtle test. Here it does seem a trifle odd to imply that the capacity of primary age children to identify the harm done to others as ‘mean’ might be a bad thing.
Returning to the altruism question, let us suppose that rules are the key, and develop a story around that. In religious households there is a clear structure and story told of reward and punishment. This might include clear identification of misdemeanours but it may also include a clear identification of entitlement to reward. It might include clear expectations of duty towards others, but modified by reflection on how to quantify and limit what counts as fulfillment of duty. The Koran has clear prescriptions about being generous to the needy – so do the Old and New Testaments. However all three sets of texts, insofar as they make positive suggestions of how to do this livably, can also be taken to imply that there is a reasonable limit to what one is expected to give. How might the adult behaviour reinforce such a perception – the way the parents talk about taxes? For Christians in the US (Catholic and Protestant) the payment of the tithe – 10%, (but ‘only’ 10% )? Religious children are not not altruistic, but perhaps they are calculatedly altruistic.
This may feed into what the report identifies as ‘moral licensing’ – the phenomenon that because I feel that I am, on the whole, righteous, I can allow myself to ease up in other areas of moral endeavour. I would certainly recognize that phenomenon in myself occasionally. It is the sort of behaviour denounced roundly in the Gospels and has proved remarkably resilient in the history of Christianity. It is worth adding that such self righteousness and the moral licensing that goes with it are not the exclusive property of religious groups. In any group that has a clear set of non-negotiable rules, those who successfully keep most of them can sometimes catch themselves taking liberties with those that are occasionally inconvenient. Nevertheless (on this account) perhaps the religious children are (at some level) telling themselves the story, ‘because I’m good I don’t need to give away that fourth sticker’.
But what of a non-religious household? Well if it is a household in the US, then it is most likely to be trans-Christian non-religious. It might have rejected the existence of God and some of the norms that some religious households down the road find very important, however, it may find inherited general norms of charity and generosity something worth holding onto – in a non-specific way. Generosity, as held up as a rule for behaviour, is thus an open-ended value rather than a defined one. If the parents have thought through their attitude to rules (this can apply to religious parents as well), they are more likely to have come out on the side of epieikeia – the prudent discernment of how to apply general rules – and they may simply have fewer of them. The child with a more dominant sense of the generosity rule and without ‘guidelines on giving’ may respond more generously than the child governed by a larger set of rules and with an intuitive understanding of prudential or reasonably-limited generosity.
Would it be different if you were a non-religious household in China or South Africa? Perhaps the stories there would have more to do with a narrative of social solidarity, common ownership and equality. Again such a parental narrative might naturally highlight an open-ended virtue of generosity. If reinforced by parental behaviour, that would be a reason for the child to respond more generously.
There is another significant and subtle set of possibilities (though further details of the conversation in the experiment might prove them invalid). Can you recall what you got anxious about as a child? Children will not only imitate the behaviour of their significant elders and repeat their wisdom, they will also want to please them. One of the key things that might be going through the heads of the children given the stickers, told they are theirs, and told about the other children is likely to be, ‘what am I expected to do?’. The question is likely to be both more intense and more complex for children coming from households where expectations are more regularly expressed and in greater number and obedience is given a higher value than autonomy. As in any area of life, both overthinking and caution can inhibit one’s spontaneity (this is certainly the case in my own experience). Is this perhaps something of what is happening here?
Let’s imagine a conversation inside the head of child A (the Samaritan) who grows up in a less introspective, less rule-bound and higher autonomy environment:
‘These are nice stickers. But this other person hasn’t got any. They can have four of mine.’
And child B, who is used to worrying about getting it right (the priest who walks by to avoid ritual pollution)…
‘He’s given me these stickers. So he wants me to have them. But he says this other person doesn’t have any. It would be nice to give that other person some of mine. Does the man want me to give them some of mine? But how many does he want me to have? Will he be upset if I give away too many? Maybe it will be alright if I give the other person three…’
These are just suggestions about possible stories we might tell to put flesh on the figures. The statistics give us a significant correlation. But they cannot tell us directly about cause and effect. These ones give us interesting material for reflection about the relationship between frameworks of rules and sanctions, the affective aspect of moral development and the nature of moral maturity. Interestingly, the challenge they raise is one that has been at the heart of self-reflective religion for millenia, and is encapsulated in the story of the Good Samaritan or the tirades against the Pharisees and Lawyers in the Gospels.
So at this point, it is appropriate to go back to a crucial assumption embedded in the report, about what counts as ‘religious’. We find it in that discussion about judgmentalism and punishment.
“Research indicates that religiousness is directly related to increased intolerance for and punitive attitudes toward interpersonal offenses, including the probability of supporting harsh penalties. For instance, within Christianity, fundamentalists tend to be more punitive and advocate for harsher corrections than non-fundamentalists”
“While this association is documented in adults of the major world religions, here the relation between greater religiousness and preference for more severe punishment is observed in development, when morality is in a sensitive and fragile period, subject to social learning and cultural practices”
The implication here seems to be that the more fundamentalist (or rule bound) you are in your religious practice, the more religious you are. This excludes, a priori, the notion that being very religious can contribute to mature (and socially useful) moral development. However, there are many forms of religion and philosophy that are not rule-bound in this sense, and which can help people grow in and chart a subtle and responsible pathway through a complicated world, in which judgment is balanced with a mercy often conspicuously absent in the wider social domain. Just because they are not fundamentalist does not make them less religious. ‘Religious’, as indicated above, can mean a lot of different things and include a variety of forms of commitment to core beliefs and group membership.
The questions raised here really focus on the relationship between the ‘natural’ capacity for altruistic behaviour and the pedagogical framework and narrative within which a child is reared. Clearly there seems to be something significant to do with the articulation of rules for behaviour within households, but what exactly we are being told requires the sort of fine analysis and qualitative story-telling that yields no simple rules of cause and effect. The report usefully challenges the easy assumption that being religious makes you more altruistic than the people down the road but it might be a bit hasty to replace that with an equally easy opposite.