Tuesday, April 13, 2010

The Equation to Surviving Andersonville (Hint: Pr (S=1) = Φ(ß(1)F + ß(2)I)

A recent scientific article attracted my attention:

Costa DL and Kahn ME, "Health, wartime stress, and unit cohesion: evidence from Union Army veterans," Demography, Feb 2010 Feb, Vol. 47, No. 1, pp. 45-66 (abstract here; PDF here)

I contacted Dr. Dora Costa at UCLA for a copy of the article (which she kindly sent) and I am reading it now.

She also sent along another article which is especially interesting:

Costa DL and Kahn ME, "Surviving Andersonville: The Benefits of Social Networks in POW Camps," American Economic Review, 2007, 97(4): 1467-1487. (PDF here)

Readers are (gently) warned that the papers are heavy in mathematics and statistics, but the authors' arguments are fascinating, especially for the Andersonville paper. While there is obviously no true "equation" to explain survival, the heavy "ciphering" in the papers goes with the rigors of social science. Yet, the introduction, discussion of results, and conclusion of both papers are very readable.

As for the Andersonville paper: briefly, they conclude that being part of a "social network" (I don't mean Facebook...see here for a scientific definition) was a key factor in surviving the prison by analyzing a mountain of data from Union army service and pension records to identify variables that had a statistically significant impact on survival. It turns out, according to Drs. Costa and Kahn, that "In two independent datasets we found that friends had a statistically significant positive effect on survival probabilities, and that the closer the ties between friends as measured by such identifiers as ethnicity, kinship, and the same hometown, the bigger the effect."

The article crosses a number of economic and social science disciplines including aging studies, game theory, and others.

Look for an interview with Dr. Costa about her interesting research and conclusions in my "Medical Department" column in The Civil War News in the coming months!


hjs21 said...

Costa & Kahn presented a lot of this stuff in "Heroes & Cowards: The Social Face of War" which I read last year. I'm not sure they made a convincing or cogent argument for whatever it was they were trying to demonstrate, but it was an interesting effort. And yes, they write like economists. ZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZ

Jim Schmidt said...

Harry - Thanks...I hadn't read "Heroes and Cowards" yet, and hadn't made the connection with the authors of these papers and the book, until you mentioned it. I may read it now. Likr I said, the papers are certainly heavy on statistics, etc., but if you dig down to the actual analysis/conclusions, it's at least plausible that unit cohesion, etc., has an impact. To be fair, in the paper, they also describe the negative and null impacts that unit cohesion can have. Anyway, I thought it was kind of like "Beautiful Mind" meets "Andersonville." Thanks for visiting! Jim