Does religious belief make you more moral: a case study in misusing data?

Initial note

This is unfortunately a sometimes negative post, and slightly longer than usual. I’m sorry about that. I have tried to be fair and positive, but I think it was important to address this issue.

Both christians and non-believers seem to want to prove that their belief makes for a better society and the opposing viewpoint is harmful. And so both sides look to research to bolster their conclusions. Trouble is, the research isn’t always unanimous, and sometimes it can be downright misleading.

As a case in point, Gregory Paul has published several papers in academic journals that claim to show religious belief leads to low prosociality. But the often-quoted claims are not backed up by rigorous analysis (as we shall see).

So what is the current academic consensus?

Definitions of ‘religion’ and morality

Some studies use the term ‘religion’ loosely while others are more precise. The following need to be distinguished:

• Religious practice, or extrinsic religion, refers to taking part in rituals and events, regardless of whether one believes in them or attends for social or other reasons. Researchers believe some of the moral benefits, and drawbacks, of religion arise from its social aspects.
• Personal belief, or intrinsic religion, focuses more on self and God. Many such believers also participate in community religious rituals (like attending church) but many do not. Intrinsic religion seems to almost always lead to good moral outcomes.
• Finally there are people who have a less formed belief – it may be classified as spirituality or quest religiosity – which may result in different moral behaviours again.

Obviously the particular definitions used and the questions asked could lead to significant differences in the assessment of religiosity.

Morals or ethics can vary across cultures and can be difficult to define, so researchers often use the term prosocial to describe behaviour that benefits the community and other people generally. Prosociality is clearly related to morality, though isn’t the same.

Religion by country

One way to examine this question is using whole country data. The largest dataset is probably that of the World Values Survey (WVS), but there are significant European and American datasets also. Analysing this data shows some broad groupings:

1. Western Europe (together with ‘western’ nations such as Canada, Japan and Australia), which have low levels of religiosity, but high wealth and generally high levels of prosociality.
2. Developing countries or third world countries which have high levels of religious belief, lower wealth and lower prosociality.
3. The United States, which is an outlier on many graphs, because it has high levels of religiosity and high wealth, but lower prosociality.

Paul agrees with most other researchers with this outline. But Paul goes on to conclude that “conservative religious ideology apparently contributes to societal dysfunction”. THis alleged causation may look plausible at first, but isn’t supported by most experts.

Causation?

The problem with the simple answer is that it leaves out too many factors and assumes causality. Which way does causation go? Is religion the only factor in prosociality? Clearly not. So how much of a factor is it?

Professional researchers use survey and statistical methods to test which factors are more likely to be causes and which ones are results of other causes. There is still much disagreement about all this, but here is what seems like the most plausible idea so far (based in part on several articles by Connor Wood and Jonathon Morgan from Boston University plus some information from WVS).

1. The most basic factors are wealth (including wealth equality) and the degree to which a country provides a secure ‘safety net’ for the most disadvantaged.
2. Inequality of wealth is one of the major causes of crime and antisocial behaviour, which effect the stability of society and reduce measures of prosociality. Thus most western countries have relatively stable and peaceful societies with high prosociality and happiness, whereas many developing countries are not as stable or prosocial. The USA is anomalous among wealthy countries in that it has high wealth inequality and high crime rate (e.g. high homicide rate) and thus lower prosociality.
3. Researchers tend to believe that religion is strong when people feel less secure (the reverse of the causation that Paul suggests). Thus religious belief has declined in western countries except USA, but is still strong elsewhere.
4. However religion is the main source of meaning, so a sense of purpose in life is reduced in western countries, resulting (for example) in higher suicide rates.

Thus it seems that, nationally, religion is not a cause of low prosociality as Paul suggests, but both are a result of different levels of wealth and security.

How did Paul get it wrong?

As far as I can see, Paul’s analysis:

• didn’t consider all the possible interactions between different factors as other researchers do to isolate and identify the fundamental correlations;
• didn’t consider the different forms of religiosity (i.e. intrinsic vs extrinsic);
• was based on only 17 highly developed countries (12 in western Europe) rather than a full world dataset;
• claimed a causal link without demonstrating it (and other people may have demonstrated the opposite link).

Many reviewers, especially of his 2005 paper, indicated that selecting different countries, measures or methods of analysis would give different results.

The big problem with national studies

Studies based on national data tend to average out the results too much to draw the sorts of conclusions Paul draws. Some countries are relatively homogeneous regarding wealth and religiosity, but others are not (e.g. China, Singapore, South Africa, India, where christians are atypical socioeconomically). Nation-based studies ignore the often large differences within a country. To assess whether religion leads to greater prosociality requires a different method.

Studies of individuals

Perhaps the largest number of studies use data that is broken down within each country to see how different segments of the community behave. These studies give a wide variety of results (often depending on what measure of prosociality is used) – some show no relationship between religious belief and prosociality, while others do; some show different responses from extrinsic, intrinsic and quest religiosity. Some examples:

• Religion influences ethics more in low-religiosity countries, and countries where there is freedom of religion.
• Extrinsic religion tends to lead to greater prejudice, but (intrinsic) belief in God tends to make people less prejudiced.
• “Frequent churchgoers are more active in volunteer work and a devout national context has an additional positive effect. However, the difference between secular and religious people is substantially smaller in devout countries than in secular countries.”
• In this study where people self reported moral responses, “Religious and nonreligious participants did not differ in the likelihood or quality of committed moral and immoral acts.”
• Studies summarised here show that heartfelt religious belief encourages compassionate behaviour, but external religious observance is less likely to.
• This analysis of 21 separate studies, mainly of European christians, found that religious faith tended to promote more traditional moral values and prosociality, but not ones associated with hedonism or change.

Is there a consensus?

There appears to be a broad consensus that religious belief is associated with greater prosociality. The following assessments of the current research indicate this:

Survey by atheist researchers

Lilienfeld and Ammirati are both researchers at Emory University and both self-confessed atheists. They did a wide ranging literature review, reported in the Sceptical Inquirer to find the current consensus, and they said: “more recent studies, as well as meta-analyses (quantitative syntheses) of the literature, have converged on a consistent conclusion: belief in God bears a statistically significant, albeit relatively weak, association with lower levels of criminal and antisocial behavior, including physical aggression toward others”.

Their final conclusion: “the data consistently point to a negative association between religiosity and criminal behavior and a positive association between religiosity and prosocial behavior. Both relations are modest in magnitude and ambiguous with respect to causation. At the same time, they cannot be ignored by partisans on either side of the discussion.”

Science on Religion blog (Boston University)

The Science on Religion blog regularly reports on religion and prosociality. While a diversity of conclusions is noted, the blog seems to support the consensus of a positive link with these words:

“a large body of research linking religiosity to prosocial behavior”
“Religion ….. does actually seem to make people more altruistic and generous.”
“religiosity ….. correlates with generosity”
“plenty of research has shown that religious people do tend to give more to charity and to volunteer more”

Review by University of Illinois

A review of a number of studies concluded: “religion shows some positive effect on prosocial behavior, but these effects are not always straightforward.”

Other summaries

Many researchers indicate that a positive religion-prosociality link is the generally accepted view and then go on to explain why they think that is. Others go on to challenge that consensus.

• “Many of the links of religiousness with health, well-being, and social behavior may be due to religion’s influences on self-control or self-regulation.” (Psychological Bulletin)
• “Numerous authors have suggested that religious belief has a positive association, possibly causal, with prosocial behavior. This article critiques evidence regarding this “religious prosociality” hypothesis” (Psychological Bulletin)
• “Recent studies have found that activating religious cognition by priming techniques can enhance prosocial behavior, arguably because religious concepts carry prosocial associations.” (Preston & Ritter)
• Ara Norenzayan of the University of British Columbia is one of the most eminent academics in the field. A 2008 paper The Origin and Evolution of Religious Prosociality said “Although sociological surveys reveal an association between self-reports of religiosity and prosociality, experiments measuring religiosity and actual prosocial behavior suggest that ….. [explaining why they think this occurs]”
  Norenzayan also said in a 2014 paper Does religion make people moral? that “mechanisms found in religions encourage prosociality towards strangers, and in that regard, religions have come to influence moral behavior …. [but] religion is not necessary for morality”

Conclusion

It seems clear then that there is a definite, but small and variable, connection between religious belief and prosociality – especially when that belief is personal and heartfelt, and less so when it is extrinsic.

This doesn’t mean that non-believers aren’t moral, or that believers are more moral than non-believers. It simply means that religious belief slightly increases the probability of a person making prosocial choices.

Neither does this finding mean that religious belief is true, though it is consistent with that. But I think it is a challenge to both believers and unbelievers. Sceptics should be more circumspect in their negative claims about the effects of belief, while christians could see this as a challenge to do better.

What none of the studies measure

All studies I have seen record people’s choices at the time of the study – they make no assessment of how a person behaved in the past, and hence don’t measure if poorly behaved people improve if they change belief. This seems to me to be a significant drawback to answering this question. Longitudinal studies (over many years) of how changes in belief (either way) led to changes in behaviour (or not) would seem to be the best way of testing this.

Addendum: finding what you want to find?

I have left this comment to the end, so that the evidence could be assessed on its merits. But it is worth noting that Gregory Paul is not qualified in any field relevant to these studies, and writes from a polemical atheist perspective (as revealed by titles of his papers and the way they are written, both of which lack the objectivity one expects in scientific papers). His methods have been questioned as unprofessional and selective – whether deliberately so I cannot say. It is therefore no surprise that some of his conclusions seem out of step with the consensus.

Some, though not all, of his critics are christians, so their responses may also be unsurprising.

I think it is unfortunate that his papers are quoted by many atheists, without mentioning the vast array of research on this topic, and apparently without any awareness that they are not representative of the field as a whole.

Read more

• Gregory Paul’s papers: Cross-National Correlations of Quantifiable Societal Health with Popular Religiosity and Secularism in the Prosperous Democracies: A First Look (2005) and The Chronic Dependence of Popular Religiosity upon Dysfunctional Psychosociological Conditions (2009).
• Critiques of Paul’s 2005 paper: NBC News, John Sexton (2005), John Sexton (2007), George Gallup and Tom Gilson.
• Positive references to Paul’s papers: Jerry Coyne, Matthew Provonsha and Zack Ford.
• A questioning review by Sue Blackmore.

Photo: Salvation Army