Thursday, July 20, 2006

Belief in God is Bad for Society

Read it on KevinWillis.net

And they can prove it, using the magic of statistical analysis. And I do mean magic. I review Cross-National Correlations of Quantifiable Societal Health with Popular Religiosity and Secularism in the Prosperous Democracies, and find it lacking.

Last May, I wrote in regards to an email I received about an earlier article (You Christians are Stupid and We Hate You), in which the assertion was made by the writer that in addition to being anti-scientific and backwards, Christianity and religiosity caused all manner of social ills. For evidence, the writer cited the following:

In response to certain claims that you make in your article, I submit the following study, which finds correlations between religiosity and various types of social dysfunction, such as child mortality, homicide, abortions, STDs, low life expectancy, etc. All of the aforementioned social ills seem to increase as the percentage of the population that believes in God increases:

http://moses.creighton.edu/jrs/2005/2005-11.html
Which, in my humble opinion, began inauspiciously by with an immediate attempt to create false academic legitimacy, and perhaps obfuscate the agenda, with its title: Cross-National Correlations of Quantifiable Societal Health with Popular Religiosity and Secularism in the Prosperous Democracies. A more direct and honest title might have been: Religion Bad, Secularism Good.

As I noted earlier, my immediate problem, and the most intractable, is that the attempt to demonstrate a correlation between religiosity and social ills casts such a statistically broad net as to be of no objective value, at least in terms of providing meaningful data about a demonstrable relationship. A satistical demonstration that two or more factors can be related on a graph is almost certain to be a demonstration of coincidence, if not selective data mining. And this can be done, and has been, on every issue from crime in the African American community (although certainly most legitimate statisticians would not argue that skin color or ethnicity itself causes crime) to arrests during full moons. Paul Broca is famous for his statistical proofs that brain size was indicative of brain power, and that white, European males all had the biggest brains, and women and non-Europeans were inherently mentally inferior and were always going to be. While his method of gather statistics and analyzing them has now been fully debunked, most of that happened in the harsh light of history. Despite all of Broca's statistical analysis, very little he maintained was true: white European males do not consistently have larger brains, physically, than other races. Women generally have smaller brains due to their smaller size physically, but numerous different tests of scholastic aptitude have shown that this does not make men any smarter than women. Indeed, modern indications (if we viewed them in a vacuum) would be that women are smarter than men.

So one should always regard any statistical analysis that assigns moral and intellectual superiority to one group or another, based on any over-riding characteristic of that group, with deep skepticism. Replace the examples of religiosity correlating with social ills in the study with examples of ethnicity correlating with social ills, or gender correlating with social ills, and see how it sounds. Or political affiliation, and put your personal political affiliation in just to get an idea of how problematic such "analysis" really is.

Interestingly, early on the authors essentially say that their study, which implies a causal relationship between religiosity and social ills, cannot demonstrate it, so they don't try (yet). This is also their explanation for not performing regression analyses, which might well contradict the fundamental implication of the study: that lots of people believing in God is bad for society.
Regression analyses were not executed because of the high variability of degree of correlation, because potential causal factors for rates of societal function are complex, and because it is not the purpose of this initial study to definitively demonstrate a causal link between religion and social conditions.
While I think some of the conclusions are generally accurate, for example:
Correlations between popular acceptance of human evolution and belief in and worship of a creator and Bible literalism are negative (Figure 1). The least religious nation, Japan, exhibits the highest agreement with the scientific theory, the lowest level of acceptance is found in the most religious developed democracy, the U.S.
... I don't think they tell the whole story. By popular acceptance of evolution, do they mean that in regards to where or not evolution occurs at all, or speciation via evolution, or that man evolved from lower life-forms? Those are all, believe it or not, different things, and many Christians accept that evolution happens, but they don't accept that man evolved from lower animals, although the evolve sufficiently for race variation and homogenization, increases or decreases in height and size, etc. And some accept that evolution happens, that man evolved from lower life forms, but believe that God's hand guided the process from beginning to end. And this is also just one example where the language bugs me, and ought to bug anyone looking for an objective scientific analysis: "Correlations between popular acceptance of human evolution and belief in and worship of a creator and Bible literalism are negative" . . . belief in a creator is not the same as worship of a creator, and neither is the same as Bible literalism and, indeed, Bible literalism means that they aren't talking about belief in or worship of a creator, or even Old Testament-consistent belief systems like Judaism or Islam, but of Christianity specifically. Although it's not entirely clear that's what they mean.

It doesn't take them long to go from confusing to deeply flawed:
A few hundred years ago rates of homicide were astronomical in Christian Europe and the American colonies. In all secular developed democracies a centuries long-term trend has seen homicide rates drop to historical lows.
Any attempt to equate the drop of homocides over centuries to a decrease in religiosity over the same period is simply non-scientific. How many other things happen culturally over a period of 100 years? From advances in education, social spending, prison-building, law-writing, greater enforcement, economic opportunity, etc. There are hundreds if not thousands of significant societal changes that went on, many of which one would rationally assume had a much more significant effect on the rate of violent crime, than a reduction in the levels of religiosity in a given country.
The especially low rates in the more Catholic European states are statistical noise due to yearly fluctuations incidental to this sample, and are not consistently present in other similar tabulations.
Translation: where the data disagrees with our pre-determined conclusions, we blame the data. However, where it agrees with our conclusions, the data is fine. A clear indication that the yard stick for the accuracy of the data is how neatly it agrees with the authors conclusions (that they admit they can't actually demonstrate a causal relationship for!).
Despite a significant decline from a recent peak in the 1980s, the U.S. is the only prosperous democracy that retains high homicide rates, making it a strong outlier in this regard.
Which might be affected by things like a gun-culture and hip-hop culture that have a limited relationship to religiosity (or even a negative correlation which is, no doubt, why that aspect is not explored in the study).

Something else striking in it's absence, given the first assertion about murder rates hundreds of years ago versus now, the change in the homocide rate, per capita, in America from 1800 to 2000, is not touched on. Likely, this is because, thanks to urbanization and increased law enforcement and increase prosperity, the homicide rate went down even as religiosity did not.
Mass student murders in schools are rare, and have subsided somewhat since the 1990s, but the U.S. has experienced many more (National School Safety Center) than all the secular developed democracies combined.
What the hell is that? And why doesn't it mention that, when known, some of the most notable school shootings (ala Columbine) were performed by irreligious kids in families that did not attend, or attended very little church. Are we to understand that it was the fact that lots of other kids attended church that made the Columbine students kill so many of their classmates? Oh, probably not, because, as the authors admit at the outset, they don't actually mean to say there is a causal relationship between societal dysfunction and religiosity. They just want to cherry-pick examples to sure make it feel like there is.
Other prosperous democracies do not significantly exceed the U.S. in rates of nonviolent and in non-lethal violent crime, and are often lower in this regard. The
This is another semantic issue for me again, one that you might miss if you blinked. But there is no where in the study where that kind of language is used in reverse--that is, nothing that says, in effect, "The US does not exceed other prosperous democracies in regards to societal dysfunction X, and is often lower in this regard." Yet here a statistic that would tend to indicate that even violent crime is often higher in secular democracies (ones that, presumably, might have much stricter gun laws and a much lesser gun culture, just to examine to factors that might impact lethality of violent crimes), and they pish-posh it away with "often lower in other places". What this means, translated, is that for crime, overall, there is no demonstrable relationship, causal or others, to even their limited definitions of religiosity, and that it goes up and down without respect to the religiosity of a given country.
Although the late twentieth century STD epidemic has been curtailed in all prosperous democracies, rates of adolescent gonorrhea infection remain six to three hundred times higher in the U.S. than in less theistic, pro-evolution secular developed democracies.
Which is to say, it has gone down significantly in the US, despite the increasing levels of religiosity over the same time period, just not nearly as much in Europe. There are almost certainly dozens of factors involved in the disparity, of which religiosity is just one. And there are almost certainly reasons why the religious beliefs of the socially dysfunctional groups are not reported on. The first is, no doubt, the lack of availability of data. But the second is likely that the data would show no correspondence, or negative correspondence, as religious kids in religious families do tend to have sex later, have fewer partners, and, believe are not, are often very careful in regards to contraception--as they would rather die that have their parents find out their sinning in the backseat on Saturday night.
At all ages levels are higher in the U.S., albeit by less dramatic amounts. The U.S. also suffers from uniquely high adolescent and adult syphilis infection rates, which are starting to rise again as the microbe’s resistance increases. The two main curable STDs have been nearly eliminated in strongly secular Scandinavia.
So one is left to conclude that church-going, God-believing religious folks are swimming with syphyllis? Or that all these churches make secualr folks more likely to engage in unprotected sex? Or the aura of Christianity in a country makes the syphyllis microbe more robust. And while there are reasons they don't even attempt to study how many religious folks suffer these social dysfunctions in America versus the non-religious folks who do, there are also reasons they aren't breaking them down by race, income, military service and immigrant status. There are lots of other correlations, in regards to STDs, that exist. And it's not mostly white, upper-income church goers who suffer from syphyllis.
Increasing adolescent abortion rates show positive correlation with increasing belief and worship of a creator, and negative correlation with increasing non-theism and acceptance of evolution; again rates are uniquely high in the U.S.
So, if fewer Americans believed in God, presumably abortion rates would go down. A skeptic might be forgiven for asking: how, exactly, does that work? Why does believing in God make people want to have unprotected sex and get abortions? Because Christians hide info about contraception? Well, there are more gas stations that churches in America, and I can buy condoms in the the bathrooms of almost all of them. Without shame! Because religious belief makes us feel guilty and so I plan not to have sex until the last minute and then I just can't stop myself and I have to have sex and can't get to a gas station because of the Christians and . . . well, I'm just not sure there is an adequate explanation for any presumed causal correlation between such factors.
Early adolescent pregnancy and birth have dropped in the developed democracies, but rates are two to dozens of times higher in the U.S. where the decline has been more modest.
And in almost all of those populations, the overall corresponding birth rate has gone down, too. And, again, there's a reason they don't break this down by race, or by individual on public assistance: a large number of adolescent pregnancies and births happen to kids from single-parent homes, almost always single-mother homes, where the family has been on public assistance for a significant portion of the teen-mother's life. The number of teen pregancies to children attending public schools versus those attending private religious schools is significantly higher (feel free to research the data yourself) by any yardstick . . . again, that's why the analysis is limited to very broad strokes regarding very limited sets of data. Because there are much more obvious and consistent relationships to things like public assistance, public school attendance, and single-parenthood and societal dysfunction than there are to religiosity.
Broad correlations between decreasing theism and increasing pregnancy and birth are present, with Austria and especially Ireland being partial exceptions. Darroch et al. found that age of first intercourse, number of sexual partners and similar issues among teens do not exhibit wide disparity or a consistent pattern among the prosperous democracies they sampled, including the U.S.
Why, how can this be? With all the religious oppression that happens in America? Perhaps because religiosity is not a significant factor in regards to teens and sex?

Now, remember earlier in the study, with the authors said this: "...potential causal factors for rates of societal function are complex, and because it is not the purpose of this initial study to definitively demonstrate a causal link between religion and social conditions [emphasis added]."Now, you tell me, how do you reconcile that statement (unless it is simply there to avoid any accountability for their inflammatory assertions) with the following:
In general, higher rates of belief in and worship of a creator correlate with higher rates of homicide, juvenile and early adult mortality, STD infection rates, teen pregnancy, and abortion in the prosperous democracies ... The United States is almost always the most dysfunctional of the developed democracies, sometimes spectacularly so, and almost always scores poorly. The view of the U.S. as a 'shining city on the hill' to the rest of the world is falsified when it comes to basic measures of societal health. Youth suicide is an exception to the general trend because there is not a significant relationship between it and religious or secular factors. No democracy is known to have combined strong religiosity and popular denial of evolution with high rates of societal health. Higher rates of non-theism and acceptance of human evolution usually correlate with lower rates of dysfunction, and the least theistic nations are usually the least dysfunctional. None of the strongly secularized, pro-evolution democracies is experiencing high levels of measurable dysfunction. In some cases the highly religious U.S. is an outlier in terms of societal dysfunction from less theistic but otherwise socially comparable secular developed democracies. In other cases, the correlations are strongly graded, sometimes outstandingly so.
Just the fact there are no other factors considered except religiosity (not GDP, not race or ethnicity, not cultural diversity, not tax burdens or public assistance or marriage rates or public education or social spending or single-payer healthcare or anything else) is a huge flaw, in regards to attempting any kind of meanginful study, in my opinion. That so many other potential ills are left out--specific rates for rape and child abuse or abandonment, let's say, or adult suicide for those not terminally ill, rates of cancer and heart disease, longevity and so on, is telling. That potential positive correlations are also completely left out of a discussion of factors that might correspond to religiosity, such as public charities, religious hospitals, scientific innovation, the size of the middle-class, etc, also point to more agenda than science being at work in this study.

A conservative fellow like myself could certainly be forgiven for seeing more anti-religious and anti-American bias than a love for the empiric pursuit of science at work in this particular analysis.

Again, I want to remind you of when the authors said the following: "it is not the purpose of this initial study to definitively demonstrate a causal link between religion and social conditions [emphasis added]." Remember that? When their conclusion says this, it makes one wonder if they remembered writing that. Just look:
The United States’ deep social problems are all the more disturbing because the nation enjoys exceptional per capita wealth among the major western nations. Spending on health care is much higher as a portion of the GDP and per capita, by a factor of a third to two or more, than in any other developed democracy.

The U.S. is therefore the least efficient western nation in terms of converting wealth into cultural and physical health. Understanding the reasons for this failure is urgent, and doing so requires considering the degree to which cause versus effect is responsible for the observed correlations between social conditions and religiosity versus secularism.

It is therefore hoped that this initial look at a subject of pressing importance will inspire more extensive research on the subject. Pressing questions include the reasons, whether theistic or non-theistic, that the exceptionally wealthy U.S. is so inefficient that it is experiencing a much higher degree of societal distress than are less religious, less wealthy prosperous democracies.

Conversely, how do the latter achieve superior societal health while having little in the way of the religious values or institutions? There is evidence that within the U.S. strong disparities in religious belief versus acceptance of evolution are correlated with similarly varying rates of societal dysfunction, the strongly theistic, anti-evolution south and mid-west having markedly worse homicide, mortality, STD, youth pregnancy, marital and related problems than the northeast where societal conditions, secularization, and acceptance of evolution approach European norms.

It is the responsibility of the research community to address controversial issues and provide the information that the citizens of democracies need to chart their future courses.
Got that? They apparently know that America is wealthier than any other nation in the world, but see's no correlation between that bit of American exceptionalism and religiosity--the idea doesn't even come up, which should speak volumes about the agenda of the authors.

You'll also notice that religiosity suddenly takes a back seat to the central conceit of study: that America is, one the whole, inferior to Europe and Japan in almost every respect, and that we need to get cracking and emulate Europe and Japan.

Oh, and wait, we're not done . . . the backwards rednecks of the south and mid-west (you know, Bush voters, Republicans, church-goers, and other extra-chromosome defectives) are where all the bad things and violence come from in America, as it turns out, while the superior, largely secular bastions of liberalism in the northeast approach "European norms". So, while the agenda is ostensibly anti-religious, one might also see it as a bit of anti-rural Europhilia, and certainly disdain for the red states there, too.

While the authors urge more research, presumably similarly agenda-laden, I don't think they do anybody any favors by taking the approach they have. What's the racial and ethnic makeup of the south and mid-west compared to the northeast? Heck, what are the population levels, the average age of the population, and how many people move into the state versus those moving out? What are the foreign immigrant levels? How much public assistance is doled out? What are the gun laws? What are the sentencing guidelines for judges? And so on and so forth. There are real reasons for societal dysfunction, and there is plenty of societal dysfunction everywhere, and wasting time and money attempting to enshrine religious bigotry with the trappings of legitimate science does not address them. Such studies, that don't include an analysis of race, gender, age, immigrant status, income level, levels of public assistance, and so on, are simply not serious and do absolutely nothing of value in regards to addressing the societal ills they proclaim to be so important.

It is also worth noting that the soup kitchens in this country are not staffed by academics. Charity hospitals are not run being run by researchers. There may be a First Secular Humanist Charity Hospital somewhere, but I haven't seen it. Most foodbanks are not run by dedicate atheists and the first responders, outside of the fire, police, and American military, to most natural and even man-made disasters tend to be Christian and Jewish organizations. The amount of good done by contemporary Christians could fill volumes that could then fill a library, and the record of secular humanist researches is not nearly so impressive. Although they do have an impressive ability to laud themselves and pat themselves on the back while simultaneously pointing their fingers, I'll find their arguments more credible when they start coming from the soup kitchen they've been serving homeless folks meals in for the past decade.

Other excellent commentary, supportive and critical, on the study can be found by following these links:

Evidence for the effects of religion on society (idea)

Magic Statistics: From Our Bulging How Not Do Do It File

Dogma Bites Man

Distant Correlations

IS RELIGIOSITY SOCIALLY BENEFICIAL?

Atheism or Christianity: Whose Fruit is Sweeter

To the church, he’s public enemy No. 1

Religion: Harmful for Society? New Study Says Yes, But Its Argument Shows Flaws

Experts say religion study is sound, but ...

Time to Choose: God or Country

No comments: