ID Critic Gregory Paul published a "study" purporting to show a strong correlation between relgious faith and societal dysfunction. The study has already been sufficiently debunked by statistician Scott Gilbreath. But I thought I would have some fun with this also.

Mr. Paul's whole study is built around the assumption that he can compare various Western nations as if religion is the only significant variable that distinguishes them. But as Gilbreath notes:

Religious faith and acceptance of evolution were the only variables compared with the incidence of social problems. This decision was justified thusly: The cultural and economic similarity of the developing democracies minimizes the variability of factors outside those being examined. This claim is highly debatable. There are many socio-economic data series that vary widely across the eighteen countries and that plausibly have a significant impact on social conditions, e.g., income distribution, proportion of GDP spent through government, social and cultural cohesion, fertility and mortality rates, age structure of the population, etc., etc. Failure to look at these and other exogenous data would introduce bias into the results, further calling them into question.

Indeed. And I would consider this a fatal flaw in the experimental design. What's worse is that Mr. Paul makes no effort to defend or substantiate his claim. He merely asserts it without supplying any references.

Mr. Paul himself seems to recognize there are some significant differences between nations. For example, when considering suicide rates, Mr. Paul only considers suicide among people age 15-24. He justifies this selection "to avoid cultural issues related to age and terminal illness." One thus wonders what other "cultural issues" have been neglected in Mr. Paul's broad-brushed approach.

Intuitively, there is good reason to suspect significant differences between the United States and many of the countries used in the comparison. This is because the USA is a relatively young country founded largely from immigrants that migrated from the countries used in the study. To borrow from Mr. Paul's simplistic approach, we could generalize and argue that the USA was recently populated by peoples looking for a place to freely practice their religion, people looking for wealth and opportunity, and people escaping the laws of their original countries. Since it is unlikely that these immigrants represent a truly random sampling from the original countries, we cannot simply assume the "cultural and economic similarity" sufficiently minimize the "outside factors."

The knock-down blow to Mr. Paul's study is his strange and flawed excuses for not applying statistical analyses to his data. Since this is nicely explained by Mr. Gilbreath, we'll skip over that. I could also cite the fact that Mr. Paul ignores a rather large body of peer-reviewed literature that addresses these issues and fails to turn up data that go his way. Mr. Paul waves such studies away because their sample sizes are not nearly as large as the data he is looking at. Yet the advantage to these smaller sized studies is that they much better control for the type of variables Mr. Paul dismisses and they are statistically analyzed and peer-reviewed.

Let's briefly consider some of Mr. Paul's correlations. Figure 2 is used to justify the claim that higher religiosity is correlated with high rates of homicide. But I don't see any trend. The USA is an "outlier" and outliers don't establish a correlative trend. Yet it's such a trend that Mr. Paul needs. Figure 3 addresses suicide rates among a small slice of the population and does not show any trend. But if you go here , you'll find that higher suicide rates seem to trend with a decrease in theism. Suicide rates would be powerful reflections of societal health, as the extreme act of taking one's life often reflects the failure of one's social surroundings. Figures 4 and 5 deal with health. The USA lags in the two measures that Mr. Paul uses, but there is no reason to think such small differences are significant or have anything to do with religion. In the USA, it is well-known that excessive inactivity (due to too much television watching and driving) coupled with excessive consumption of fast foods contributes significantly to bad health. Mr. Paul makes no effort to determine if factors such as these are in play. In fact, many studies show a positive relationship between health and relgion. For example, according to one review that just came out: "A large proportion of published empirical data suggest that religious commitment shows positive associations with better mental and physical health outcomes. There are relatively few studies showing no effect or negative effect of religiosity on health outcomes. " (Aukst-Margetic B, Margetic B. 2005. Religiosity and health outcomes: review of literature. Coll Antropol. 29(1):365-71.) This raises the question of whether health in the USA would be much worse if relgion was removed.

In figure 8, Mr. Paul shows a seeming correlation between high abortion rates and religion. But here, the comparison is invalid, as the USA has among the least restrictive abortion laws in the world. For example, a comparison of such laws shows that, except for Canada, all other countries used by Mr. Paul have restrictions that include parental authorization, a ban beyond 12- or 18-months gestational age, and/or required justifications in socioeconomic or mental health terms (Rahman A, Katzive L, Henshaw SK. 1998. A global review of laws on induced abortion, 1985-1997. Int Fam Plann Persp. 1998 Jun;24(2):56-64). For example, compared to US standards, the English tightly regulate abortion:

Technically the law did not legalise abortions, but rather provided a legal defence for those carrying them out. Abortions can legally be performed under certain conditions – the first is that continuing with the pregnancy involves a greater risk to the physical or mental health of the woman, or her existing children, than having a termination. The woman's "actual or reasonably foreseeable future environment" may be taken into account. Abortion up to 24 weeks is also allowed if there is a substantial risk that the child when born would suffer "such physical or mental abnormalities as to be seriously handicapped". An abortion must be agreed by two doctors (or one in an emergency) and carried out by a doctor in a government-approved hospital or clinic.

Mr. Paul's data indicates that more restrictive abortion laws do succeed in decreasing the incidence of abortion.

Thus, even taken at face value, most of Mr. Paul's correlations fall apart.

Finally, just for fun, let me mimic Mr. Paul's ham-handed approach. I decided to look at a subset of the countries that he mentions to better control for the variables he assumes to be non-existent. I dropped Japan from the analysis because Japan clearly has a very different cultural history when compared to the USA and European nations. I then decided to look at Germany, France, and England, as these are the Big Three in Europe whose status most closely mimics the USA. I included Canada and Australia, because these English-speaking countries are perhaps most like the USA, themselves being founded recently by immigration. Finally, I added Sweden and Denmark into the analysis, since these would represent the most secular countries in this analysis. I then used Mr. Paul's Figure 4 to eyeball the percentage of the populations that strongly believe in God and consulted the UN's Human Development Report 2005 for other data.

Mr. Paul's decision to investigate crime by focusing on homicides seemed awfully arbitrary. So I looked at the total crime rate as shown in Figure A and one eyeballs a weak correlation between less crime and a higher rate of belief in God.

Figure A

The correlation looks even stronger if we focus only on sexual assault (Figure B).

FigureB

At least this trend is supported by a truly scientific analysis by sociologists Colin Baier and Bradley Wright, who, in conducting a meta-analysis of over 60 studies, showed "that religious beliefs and behaviors exert a moderate deterrent effect on individuals' criminal behavior" (Baier, CJ and Wright, BRE. 2001. "˜If you love me, keep my commandments": A meta-analysis of the effect of relgion on crime. J. Res. Crime and Delinquency 38:3-21).

Figure C shows a modest trend that relates higher unemployment rates with secularism.

FigureC

In fact, figure D shows what looks like a much stronger relationship between high rates of long-term unemployment and secularism.

FigureD

It's even more astonishing when we look at the GDP per capita (figure E), the traditional way of measuring standard of living. A clear trend between acquiring wealth and religious faith is seen.

FigureE

Finally, Figure F shows population growth, where secularism appears to be associated with societies that are bordering on negative growth.

FigureF

Conclusion: Taken together, these data show a disturbing correlation between loss of belief in God and societal dysfunction. Chronically high unemployment, along with long-term unemployment (which may signal laziness), high rates of crime, including sexual assault, high suicide rates, and a lower standard of living seem to go hand in hand with rejection of theism. What's troubling is that these may be the symptoms of dying societies, as evidenced by the remarkably lower fecundity of the more secular nations.

Oh, and just to remind you all, I'm just copying Mr. Paul. The only difference between my "study" and Mr. Paul's is that I recognize this does not belong in the literature. And no news media will pick up on it.

In the next blog, we'll have to do a "post-mortem."

UPDATE: A New Finding

This entry was posted on Friday, September 30th, 2005 at 12:26 am and is filed under The Debate. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site. The trackback link is: http://telicthoughts.com/old-study-meet-new-study/trackback/