Slashdot Mirror
Statistical Analysis Raises Civil War Death Count By 20%
Hugh Pickens writes "For more than a century, it has been accepted that about 620,000 Americans died in the the bloodiest, most devastating conflict in American history. But now, BBC reports that historian J. David Hacker has used sophisticated statistical software to determine the war's death toll and found that civil war dead may have been undercounted by as many as 130,000. Hacker began by taking digitized samples from the decennial census counts taken from 1850-1880. Using statistical package SPSS, Hacker counted the number of native-born white men of military age in 1860 and determined how many of that group were still alive in 1870 and compared that survival rate with the survival rates of the men of the same ages from 1850-1860, and from 1870-1880 — the 10-year census periods before and after the Civil War. The calculations yielded the number of 'excess' deaths of military-age men between 1860-1870 — the number who died in the war or in the five subsequent years from causes related to the war. Hacker's findings, published in the December 2011 issue of Civil War History, have been endorsed by some of the leading historians of the conflict. 'The difference between the two estimates is large enough to change the way we look at the war,' writes Hacker. 'The war touched more lives and communities more deeply than we thought, and thus shaped the course of the ensuing decades of American history in ways we have not yet fully grasped. True, the war was terrible in either case. But just how terrible, and just how extensive its consequences, can only be known when we have a better count of the Civil War dead.'"
Shitloads died.
Let's focus on why it happened and how to prevent things like it to happen again.
Like, for instance, the government not being hellbent on forcing it's will above the states.
Not even written like an american, but as a person who saw what government trying to force it's will down upon unwilling states resulted in.
And, yeah, the same thing is happening right now, both in Europe and in USA.
I think it would be fairer to say systematically misunderstood. As soon as any project becomes even moderately complex, understanding causes of loss can become difficult.
[FUCK BETA]
The approach was a pretty sound one; take census counts before and afterwards, calculate the expected joiners (through aging and immigration) and leavers (through death, aging, and emigration), and then compare with the second count. We have enough knowledge to do this pretty accurately for a large population, so the death estimates should be pretty tight.
[FUCK BETA]
I actually dislike this kind of numeric analysis. I don't think it is appropriate for history. So there are 'missing people' from the 1860s...those missing people could easily have gone to Canada or Mexico. They could have emigrated to Europe. They could have headed out West and been out of touch with US authorities for years at a time, missing censuses and the like. They certainly had motivation to flee...there was a huge war with drafts on both sides going on, why not head out?
This study is certainly using census data, with all of its warts and flaws.
HBI's Law: Frequency of calling others Nazis is directly correlated with the likelihood of the accuser being Communist.
Well, you are entitled to your opinion. But as the analysis did its best to allow for immigration and emigration, and did comparisons with other periods, your argument needs a bit more beef than "dislike". In my own field of interest, naval history, things like manning levels, pay rates, construction, weights of shot, water capacity, access to navigational equipment and the like turn out to be far better predictors of outcomes than the "Great Man" ideas of historians of the past. The outcome of the Battle of Britain was almost entirely determined by engineering factors - radar, and the fact that the British fighters were developed a little later than the German ones and so benefited from better engineering. The massacres in Rwanda and El Salvador can be better understood by analysing population density, land use and economic power than by speculation over tribal or political conflict. Proper statistical analysis of history - not numeric analysis, whatever that is - is not only illuminating in itself but could eventually give models with predictive power.
From scarped cliff or quarried stone she cries "A thousand types are gone, I care for nothing, no not one."
I think it was Jay Gould (and if not I apologise to his shade) who observed that the usual distinction between "hard" and "soft" science is completely backward. Physics advanced faster than chemistry, and chemistry faster than biology, because in fact it is physics that is "not-hard", and especially the social sciences and economics that are very hard indeed. We had a model of dynamics that was "good enough" by 1700AD, but the causes of infectious diseases wouldn't be understood for roughly another 150 years. (We've just seen what happens when a load of bankers think physicists are capable of writing financial models).
However, as the originators of CERN would probably agree, because something is difficult is not a reason not to try it.
In my own experience, most of the people who object to statistical modeling do so because the results confound cherished beliefs. For instance, the arts graduates who run the British Home Office despise statistics because so many studies have shown that their approaches to crime don't work, and don't want to know about medical and psychological studies of the effects of various drugs because the results do not suit the agenda of the Daily Mail and the drinks companies. During WW2, the High Command of the RAF had a statistical wing that was demonstrating that (a) carpet bombing was a failure and (b) air crews did not become safer with experience. So what did they do? Ignored them, of course. But that is all the more reason why they should be done, and done as rigorously as possible.
From scarped cliff or quarried stone she cries "A thousand types are gone, I care for nothing, no not one."