Recently, a study was brought to my attention. This study was an examination of correlations between religiosity (belief in God, church attendance etc) and quantifiable metrics of social health (rates of homicide, suicide, teenage pregnancies, abortion, STDs etc). The paper to which I refer is published in the Journal of Religion and Society, a refereed journal published through the Kripke Center at a Jesuit Catholic school, and is written very objectively, appearing to be free of any ‘secular’ agenda that the uninitiated might be tempted to suggest.
At the end of the manuscript, a number of figures are shown, plotting metrics of religiosity against those of social health. Some of these are largely in line with what one might expect, but some show trends that might come as a significant shock. I’d like to draw your attention to (in my opinion) some of the more startling correlations, in the hope of understanding the meaning behind them.
Figure 1 shows correlations between acceptance of human evolution and various metrics of religiosity. In a certain manner, acceptance of human evolution is a metric for popular scientific literacy. A reasonable anti-correlation is seen between acceptance of evolution and all positive measures of religiosity (for example, as belief in God rises, the acceptance of evolution falls). This probably comes as no surprise.

Figure 2, which plots homicide rates against the metrics of religiosity, shows no statistically significant correlation. That is, the murder rate is completely unaffected by beliefs of a society. Likewise, Figure 3 shows no correlation between religiosity and suicide rates in young adults, and Figure 5 indicates no correlation with life expectancy.
However, Figure 4 is staggering. It shows a tremendously significant positive correlation between the mortality rates of the under-5-year-old population and religiosity: The higher the belief in God, the higher the infant mortality rate. And significantly so; a doubling of belief in God brings a 50% increase in mortality rates. Furthermore, the correlation is tightest around belief in God; more so than the attendance of religious services. For example, by comparing the plot against absolute belief and the plot against service attendance, the data for Ireland and Italy (which are outliers in the latter, but lie in close correlation in the former) indicate that the correlation really is with belief in God, rather than with church
attendance. (Note that these are two countries with strong cultural ties to the church (Roman Catholic), such that it is not uncommon for those who have tenuous personal belief to attend church services regularly).
Surprisingly, Figure 8 shows an unarguably strong positive correlation between teenage abortions and the metrics of religiosity, and Figure 9 shows positive correlation with pregnancies.
So, what are the data telling us? The inarguable fact is, in societies where a belief in God is more prevalent, there is more social dysfunction, often in the very realms to which the prevalent ‘religious’ teaching is most contrary. For example, it is interesting that in a society which is more vocally against all abortion, the abortion rates in teenage pregnancies (i.e. those which have the largest proportion of abortion for social reasons, rather than medical ‘necessity’) are highest. Further, there is no instance in any metric of a positive correlation be
tween religiosity and social health; either the metrics show increased religiosity coincides with more social problems, or there is no correlation whatsoever.
Why is this? We know from personal experience that genuine belief in God ought not induce such trends. We also know that, as C.S. Lewis maintained, the moral law is ingrained innately within us – that it is all-pervasive, present irrespective of theological stance. This is indicated by the data on homicide – people are no more or less likely to murder based on their beliefs.
It seems like the social problems for which there are the strongest correlations are biased towards those of youth: teenage pregnancies and abortions, particularly. It may be that, in those societies where there is the m
ost ‘religious’ pressure to conform to the ‘rules,’ there is a greater likelihood of the youth ‘careering off the rails.’ One can certainly see similar trends with attitudes to alcohol: those who were raised in cultures in which alcohol was an everyday substance throughout their childhood (wine with dinner etc) have a lower tendency toward binge-drinking in their teenage/early-adult years.
Alternatively, perhaps the correlation is actually in the opposite direction. One could argue that, in those societies where there are more social problems generally (for whatever reason) people are more troubled, and more likely to look to God. We all know, for example, how times of tribulation (for ourselves or out loved-ones) can invigorate our prayer life.
At the end of the manuscript, a number of figures are shown, plotting metrics of religiosity against those of social health. Some of these are largely in line with what one might expect, but some show trends that might come as a significant shock. I’d like to draw your attention to (in my opinion) some of the more startling correlations, in the hope of understanding the meaning behind them.Figure 1 shows correlations between acceptance of human evolution and various metrics of religiosity. In a certain manner, acceptance of human evolution is a metric for popular scientific literacy. A reasonable anti-correlation is seen between acceptance of evolution and all positive measures of religiosity (for example, as belief in God rises, the acceptance of evolution falls). This probably comes as no surprise.

Figure 2, which plots homicide rates against the metrics of religiosity, shows no statistically significant correlation. That is, the murder rate is completely unaffected by beliefs of a society. Likewise, Figure 3 shows no correlation between religiosity and suicide rates in young adults, and Figure 5 indicates no correlation with life expectancy.
However, Figure 4 is staggering. It shows a tremendously significant positive correlation between the mortality rates of the under-5-year-old population and religiosity: The higher the belief in God, the higher the infant mortality rate. And significantly so; a doubling of belief in God brings a 50% increase in mortality rates. Furthermore, the correlation is tightest around belief in God; more so than the attendance of religious services. For example, by comparing the plot against absolute belief and the plot against service attendance, the data for Ireland and Italy (which are outliers in the latter, but lie in close correlation in the former) indicate that the correlation really is with belief in God, rather than with church
attendance. (Note that these are two countries with strong cultural ties to the church (Roman Catholic), such that it is not uncommon for those who have tenuous personal belief to attend church services regularly).Surprisingly, Figure 8 shows an unarguably strong positive correlation between teenage abortions and the metrics of religiosity, and Figure 9 shows positive correlation with pregnancies.
So, what are the data telling us? The inarguable fact is, in societies where a belief in God is more prevalent, there is more social dysfunction, often in the very realms to which the prevalent ‘religious’ teaching is most contrary. For example, it is interesting that in a society which is more vocally against all abortion, the abortion rates in teenage pregnancies (i.e. those which have the largest proportion of abortion for social reasons, rather than medical ‘necessity’) are highest. Further, there is no instance in any metric of a positive correlation be
tween religiosity and social health; either the metrics show increased religiosity coincides with more social problems, or there is no correlation whatsoever.Why is this? We know from personal experience that genuine belief in God ought not induce such trends. We also know that, as C.S. Lewis maintained, the moral law is ingrained innately within us – that it is all-pervasive, present irrespective of theological stance. This is indicated by the data on homicide – people are no more or less likely to murder based on their beliefs.
It seems like the social problems for which there are the strongest correlations are biased towards those of youth: teenage pregnancies and abortions, particularly. It may be that, in those societies where there is the m
ost ‘religious’ pressure to conform to the ‘rules,’ there is a greater likelihood of the youth ‘careering off the rails.’ One can certainly see similar trends with attitudes to alcohol: those who were raised in cultures in which alcohol was an everyday substance throughout their childhood (wine with dinner etc) have a lower tendency toward binge-drinking in their teenage/early-adult years.Alternatively, perhaps the correlation is actually in the opposite direction. One could argue that, in those societies where there are more social problems generally (for whatever reason) people are more troubled, and more likely to look to God. We all know, for example, how times of tribulation (for ourselves or out loved-ones) can invigorate our prayer life.
So, the data show that these trends exist, but I don't yet have a convincing understanding of what cause lies behind them.
Thoughts, friends?