Religiosity and Murder – an Impromptu Statistical Analysis

“In 1962 we kicked prayer out of the schools. In 1963 we kicked God's word out of ours schools. In 1980 we kicked the Ten Commandments out of our schools. We've kicked God out of our public school system. And I think God would say to us, 'Hey, I'll be glad to protect your children, but you've got to invite me back into your world first. I'm not going to go where I'm not wanted. I am a gentleman.’” ~ Bryan Fischer, American Family Association

“We ask why there's violence in our schools, but we have systematically removed God from our schools. Should we be so surprised that schools would become a place of carnage?” ~ Mike Huckabee, Fox News Channel

***

These statements imply that religiosity is negatively correlated with violence. I usually find that the more vocal a religious proponent, the more often that person is just wrong, so I decided to look at the statistics.

Generally speaking, I think few would argue that murder is the most violent crime in our criminal repertoire. Below are analyses that explore the association between religiosity (the independent variable) and murder (the dependent variables). The data used were copied from the sources at the end of the analysis. Data were normalized by squaring or square root transformations, as necessary.

Rates of religiosity and murder for the fifty states are considered. A Pearson correlation test between religiosity and murder rates yields a statistic (r) of 0.541, indicating a strong positive association between the variables (p < .0001). That is, the more religious a state is, the higher the murder rate is likely to be.

Of course, if would be inappropriate to infer a causal relationship at this point; the analysis is overly simplistic and ignores other potential causal variables. I mean, for example, we’ve all heard about how criminal “certain” people tend to be. (You know… <whispers> … poor people. And <whispers even more quietly> … black people.) Hispanics too.

In fact, I do find strong positive correlations between murder rate, on the one hand, and these other variables: (1) income disparity (as reflected in Census Bureau Gini Coefficients) (r = 0.548, p < .0001); and (2) the proportion of African-americans in state populations (r = 0.674; p <.0001). (The reasons for this are no doubt several; they are also not the subject of this presentation. I do not, however find any more than a very weak and insignificant correlation between murder rates and the proportion of Hispancs in a population and that variable is not further considered herein.)

So, given that highly religious states (I’m looking at you, American South, primarily) tend also to be home to a plentitude of both African-americans and poor people, is it not reasonable to suggest that the correlation is misleading…that religious folks and blacks and poors just happen to reside together in multitudes in these states and are, in fact, the cause of higher murder rates? You know, blame it on the poors and the blacks.

Statistically, it is a valid question. A necessary question, in fact. Happily, we can control for the effects of these variables with the use of partial correlation analysis. That is, we can focus in on the variable of immediate interest – in this case, religiosity.

Controlling for the state-level proportions of African-americans, there is a moderate and significant (at the .05 level) positive correlation between religiosity and murder rates (r = 0.321, p = .024).

Controlling for income disparity, there is a moderate and significant positive correlation between religiosity and murder rates (r = 0.487, p = .0003).

And controlling for both the state-level proportion of African-americans and income disparity, there is a moderate and significant positive correlation between religiosity and murder rates (r = 0.342, p = .017).

Which is to say, Bryan and Mike appear to be … well … wrong. Imagine.


Anyway…


***


Sources:

http://www.gallup.com/poll/153479/Mississippi-Religious-State.aspx?utm_source=alert&utm_medium=email&utm_campaign=syndication&utm_content=morelink&utm_term=Politics%20-%20Religion%20-%20Religion%20and%20Social%20Trends%20-%20USA#1

http://www.fbi.gov/about-us/cjis/ucr/crime-in-the-u.s/2011/crime-in-the-u.s.-2011/tables/table-5).

http://www.statehealthfacts.org/comparebar.jsp?ind=6&cat=1).

http://www.census.gov/prod/2012pubs/acsbr11-02.pdf .