Thursday, July 09, 2009

No Need to Over-Complexificate This

There is no need to overcomplexificate* this. I think the following study goes a couple of steps too far.

The Heritage of Herding and Southern Homicide: Examining the Ecological Foundations of the Code of Honor Thesis

Robert Baller, Matthew Zevenbergen & Steven Messner Journal of Research in Crime and Delinquency, August 2009, Pages 275-300

Abstract:
The authors examine the ecological foundations of the thesis of a "code of honor" as an explanation for southern homicide. Specifically, they consider the effects of indicators of ethnic groups that migrated from herding economies (the Scotch-Irish), cattle and pig herding, and the relative importance of agricultural production across different areas in the Old South. Using county-level data on argument-related White male homicide offenders (1983 to 1998) from the Federal Bureau of Investigation's Supplementary Homicide Reports, the authors observe the theoretically expected positive interaction between the proxy measure of the presence of Scotch-Irish communities, namely, the percentage of churches that were Presbyterian in 1850, and the number of cattle and pigs per capita in 1850. They also find a negative effect of an index of crop production in 1850 on argument-related offending. The overall pattern of these findings is highly consistent with the herding thesis advanced by Nisbett and Cohen.


Herding economies? That's full of sheep dip. It's pretty simple: if you give a bunch of Scottish Presbyterians guns, a LOT of people are going to die. And you can call it a code of honor. But it's more like one of Angus's jokes:

What is the origin of copper wire? It was a shortage: Two Scots, one penny.

*GW Bush may never have said this. But I bet he did.

(Nod to Kevin L)

6 comments:

The Dude said...

This might be worthy of a Russ Roberts post explaining the limitations of statistics.

Tom said...

It makes me thoughtful -- here's what it made me think:

1. Freedom of speech makes it easier to spot the idiots.

2. If you set out to prove your pre-conceived notions, you will succeed.

John Thacker said...

I've read the original Nisbett and Cohen book referred to in the research. They actually do have evidence from a variety of herding societies, not just the Scots and Scotch-Irish.

They also have some nice statistics illustrating that homicide rates have generally been higher for the same size/type of cities in the South, because a small increase in stranger crime in the North is overridden by a higher increase in crimes of passion and other murders of people you know in the South.

Mungowitz said...

Well, as usual credit goes to John T here, for actually knowing what is going on.

I just reached some wild conclusions from reading an abstract. So whatever things I accused others of, I am myself more guilty. Thanks, John!

prisonrodeo said...

More generally (and IMO): Criminologists have the empirical toolkit of political scientists, but the theoretical apparatus (read: baggage) of sociologists. The consequences are things like this.

John Thacker said...

The book is interesting, and I suggest reading it.

This paper seems like an attempt to provide additional evidence for the thesis already well-explored in the book. Not an extremely ambitious peace of research, but I'm sure Nisbett is happy to have other people paying attention to his thesis. At least the authors here are pretty upfront about the inspiration for their research; they'd have to be.

Nisbett at UMich is also the guy who did some studies a while back showing that Southerners are indeed more polite when not upset, but get much more angry when they feel that others have violated politeness norms. Northerners aren't polite, but they don't expect other people to be so either.