Thursday, December 20, 2012

God and Guns

Steve Greene takes issue with claims by Mike Huckabee and others that gun violence is a result, in part, of our society having "turned away from God." Steve takes the time to actually apply some data to the question, noting that the U.S. is one of the more religious nations in the world but nonetheless has one of the highest homicide rates.

Here's another way of looking at the question. In the scatterplot below, I have plotted out the percent of Americans who claim to attend a church, mosque, or synagogue at least weekly with the rates of firearms deaths per 100,000 residents in each U.S. state. What do we find?
The relationship between religiosity and gun deaths is positive. Now, this doesn't mean that religiosity causes violence, and there are a lot of other factors out there that predict gun violence, including poverty, local gun laws, and culture (notice how the upper right quadrant is almost entirely southern states?). But it seems reasonable to conclude that our society's "turning away from God" is not a primary reason for gun deaths.

6 comments:

  1. A great day for any Political Scientist-- mention in Mischiefs of Faction! Thanks! Didn't even think of using state-level data. Damn, I love data (the bane of demagogues everywhere).

    ReplyDelete
  2. I smell ecological fallacy. Huckabee et al aren't arguing that violence in the aggregate is caused by religiosity in the aggregate. The argument is that INDIVIDUALS who grow up without the moral anchor that religious values provide can turn to violence.

    ReplyDelete
    Replies
    1. Possibly. I'm not totally sure what Huckabee's argument is. Is it that people of faith are less likely to be gunned down because God protects them? Is it that people of faith are less likely to pick up a gun in anger? Is it that societies that are faithful are protected by God? If this is an argument about society, then he's really not making an individual-level claim. But perhaps we shouldn't take him so seriously.

      Delete
  3. Fact-checking a claim about society turning away from God requires longitudinal data, not cross-sectional data. Steve Greene posted longitudinal data on homicide rates back to the 1990s, but religious conservatives do not perceive the Clinton years as the golden age of religion in America. Here is a more appropriate chart illustrating that, while the US homicide rate rose after prayer was removed from public schools, the rate has decreased just as much for some other reason: http://thepublicintellectual.org/wp-content/uploads/2011/03/Homicides-1900-2010-2.jpg. But the current US violent crime rate is more than twice the 1960 rate, so focusing on the homicide rate might be perceived as cherry-picking a metric, unless someone is claiming that religious values protect society from only a certain type of violent crime.

    ReplyDelete
  4. Please your own analysis, then. In particular, I'd like to know how to control for the increase in weapon technology, changes in availability through laws, and changes in production (demand) as controls.

    ReplyDelete
    Replies
    1. Hi. If this comment was directed at the second and/or third Dec 20 replies, I hope you agree that conducting one's own analysis after each peer review is neither obligated nor wise...Responding to the more specific request about handling control variables -- and providing a more general consideration of the proper method for the fact check -- would require resolving elements that are currently unclear, such as whether the fact check is for a specific Mike Huckabee claim or for a summary of non-identical claims made by Huckabee and others.

      Delete