Confusing Research on the Impact of Religion on Children's Altruism

A recent study by Jean Decety of the University of Chicago and his collagues sets up an experiment based on sticker-sharing and fake pushing among religious and non-religious children to arrive at a pretty hefty conclusion: “[The findings] call into question whether religion is vital for moral development, supporting the idea that the secularization of moral discourse will not reduce human kindness—in fact, it will do just the opposite.”

One might say that this conclusion is laughable, but the media reaction was disturbing. The web was bombarded with headlines claiming “Religious Kids Tend to Be Mean and Selfish Little Jerks,” “Religion Makes Children More Selfish,” and “Religious Children are Meaner than Their Secular Counterparts.” But despite these claims, rigorous social science has shown that religious practice delivers incomparable benefits to society. It is normal and healthy for academics to disagree on the impact of religious practice on different aspects of life. Over the long haul, this helps to clarify reality. However, it is an entirely different pursuit, and not an intellectually honest one, for researchers to intend to knock down religious beliefs and practice. Decety et al. may be doing the latter; their future research will tell. In this study, their handling of the known literature on religious practice and their poor method raises concern that, rather than seeking to add clarity to knowledge, they are only adding confusion.

Past research has repeatedly confirmed the overwhelmingly positive impact that frequent religious practice has on societal outcomes. The Marriage and Religion Research Institute (MARRI) has shown that 44 percent of adults who attended religious services monthly or more as an adolescent have volunteered in charitable activities within the past year, whereas 33 percent of those who attended monthly or never volunteered. Arthur Brooks, then a researcher at Syracuse University and now president of the American Enterprise Institute, conducted the Social Capital Community Benchmark Survey (SCCBS) that drew almost 30,000 observations from fifty U.S. communities, and used rigorous regression to control for political beliefs, income, education level, gender, age, race, marital status, and area of residence. Brooks found that, when all controls are applied, religious people are 23 percent more likely than their secular counterparts to donate money, and 26 percent more likely to volunteer. On average, a religious person gives $1,388 more than a secular person, and volunteers on 6.5 more occasions.

Moreover, Brooks points out that “Religious people are more generous than secular people with nonreligious causes as well as religious ones.” Religious people are 7 percent more likely than their secular counterparts to volunteer for neighborhood and civic groups, 20 percent more likely to help the poor or elderly, 26 percent more likely to volunteer in school/ youth programs, and 10 percent more likely to give to charitable causes.

Just last year a major German-Swiss study investigating issues similar to Decety’s summarized:

The question of whether religiosity is linked to prosocial behavior is currently hotly debated in psychology. This research contributes to this debate by showing that the nature of individuals’ religious orientations and their relationships to prosociality depend on their country’s social enforcement of religiosity. Our analyses of data from more than 70 countries indicate that in countries with no social pressure to follow a religion, religious individuals are more likely to endorse an intrinsic religious orientation (Study 1), engage in charity work (Study 2), disapprove of lying in their own interests (Study 3), and are less likely to engage in fraudulent behaviors (Study 4) compared with non-religious individuals. Ironically, in secular contexts, religious individuals are also more likely to condemn certain moral choices than non-religious individuals (Study 2). These effects of religiosity substantially weaken (and ultimately disappear) with increasing national levels of social enforcement of religiosity.

Let us look a bit more closely at the Decety study. Here are some of the primary mistakes in it:

1.  The reputation of religion now rests on stickers and bumping. Essentially, researchers assessed the altruism of religious and non-religious children by the children’s willingness to share stickers. Each child was presented thirty stickers and told to choose his/ her ten favorite. Next, researchers told the child that there weren’t enough stickers for all the children, and asked the child to anonymously place any stickers he/ she would be willing to share in an envelope. Christian children placed an average of 3.33 stickers, Muslims placed 3.20, and non-religious children placed 4.09.

Next, researchers measured how judgmental religious and non-religious children are by showing each child a series of dynamic scenarios in which one person is pushing or bumping another person (either purposefully or accidently), and assessing the child’s reactions. Muslim children labeled the interpersonal harm as meaner than did Christian children, and Christian children judged the actions to be meaner than non-religious children. Muslim children gave harsher ratings of punishment for the pushers, while there was no significant difference in punishment ratings between Christian and non-religious children.

The response that comes to mind: Give me a break.  That this rather simple study be flaunted to the lay public as proof of the impact of religion is an insult to the academy and to the profession of journalism. 

2. They use an unrepresentative sample. Lead researcher Jean Decety assessed 1,170 children between the ages of 5 and 12 years from six countries to represent the actions of religious children across the world. In the sample, 23.9 percent identified as Christian, 43 percent as Muslim, 27.6 percent as not religious, and 5.2 percent as either Jewish, Buddhist, Hindu, agnostic, or other (0.3 percent are unaccounted for). There are a number of problems with this sample that the authors do not address.

First: Children were selected from Chicago (United States), Toronto (Canada), Amman (Jordan), Izmir and Istanbul (Turkey), Cape Town (South Africa), and Guangzhou (China). It is peculiar, to say that least, that, in a study assessing how religious sentiments are manifested in behavior, researchers include countries that do not allow the free practice of religion. That alone skews and invalidates the results, as the German-Swiss study shows.

Second: The child samples seem to be opportunity samples (or snowball samples). The co-authors are all psychologists, and therefore accustomed to extrapolating from numbers not nationally representative, much less globally representative. The children seem to be from the cities where the different co-authors work.  That is far from nationally random and far from representative.

Third: The breakdown of religions in the sample does not reflect the world’s breakdown of religions. In the study, 23.9 percent are Christian and 43 percent are Muslim; in actuality, 31.4 percent of the world is Christian and 23.2 percent is Muslim.

Four: The sample consists of children ages 5 to 12 years old—a period of major developmental change for children, including significant changes in notions of justice. For academics, experiments help clarify the psychological aspects of altruistic behavior of children in mid-childhood. Only after years of research that control for an increasing number of variables will these experiments yield insights. 

Five:  The study broadly concludes that religion is bad for altruism.  If that conclusion were granted, an even bigger challenge remains for the authors: What is it about religious practice that, in the years between childhood and adulthood, flips the results so that suddenly religion encourages prosociality (as it does for adults)? From academics hostile to religion (and hostility to religion is overrepresented in academia) the response can be anticipated: Religion has nothing to do with people being good. This very argument seems to be the objective of Roy Sablosky’s “Does Religion Foster Generosity?”

3. The measures used seem far removed from reality. These researchers determined that generosity in our world is best understood by sticker-sharing and contrived acts of meanness. However, that fails the common-sense credulity test.  The professors need to come up with more realistic experiments.

Professor Luke Galen of the University of Nebraska has spearheaded much of the research debate on these issues, especially in his 2012 publication, “Does Religious Belief Promote Prosociality? A Critical Examination.” The conclusion of the abstract states:  “These factors necessitate a revision of the religious pro-sociality hypothesis and suggest that future research should incorporate more stringent controls in order to reach less ambiguous conclusions.”   

Galen is correct. Religious practice and teachings have an intricate impact on the everyday functioning of society, and should be further investigated. Religion has nothing to fear and everything to gain when the social sciences tease out the variables in play.