Many reader reactions to my earlier pieces this week on military death rates, and the political propaganda purposes to which such data are put, raised several objections and counter-arguments.

I appreciated the discussions and feedback, both positive and negative, and even those who found fault, did so respectfully and with information to back up their arguments. A rare circumstance, unfortunately, in my experience, and I wanted to draw attention to and return the compliment of civil disagreement and debate.

Blame this son of an Insurance Actuary – do other Insurance Industry progeny have this same obsession with statistics? – but I think it vitally important to address the statistical problems and issues that were raised in these discussions, as well as political or agenda-based concerns. Statistical means and methods first.

Several people made the same mistake I did at first, which was to mistake and misunderstand the data presented. Others, including myself and the author of the NY Sun piece, mixed two different cuts of data between the two examples, any selection of pre-Iraqi war military death rate data and what was in fact exclusively raw data on total military deaths in Iraq. What slices of data you use will always influence the result you can portray. One critic answered the NY Sun Op Ed’s admittedly arbitrary but politically convenient basis for comparison of early Clinton Years to Iraq war years, with an equally arbitrary and equally convenient Later Clinton Years to Iraq war years.

The table that was presented was of military death rates from 1980-2004, not total deaths. Why is this important? Because unless you use death rates, you’re comparing raw number of deaths between years in which there were differences in the number of military. The tables showed the raw number of deaths in various categories per 100,000 individuals. That way, the rates in the various categories can be compared across periods in which the numbers in service fluctuated.

To pluck any segments of raw data out and use for comparison, without considering the basis and assumptions for those numbers, and finding a way to reconcile and/or extrapolate, means you’re comparing apples and oranges. Working through the available data, and selecting just those portions for comparison that exaggerate the predetermined effect you are seeking, intentionally introduces selection bias. If you pick only the data that supports your intent, and ignore data that refutes it, you get what you pay for. Call it the MSM Approach to Reporting on Iraq. (Because it is.)

Several commenters pointed out that death rates attributed to accident were steadily decreasing throughout the period in question, and that higher combat death rates are therefore masked in the composite, and this is no doubt true. I’ve served in the military, as a senior non-commissioned officer (NCO), and virtually my entire career involved in training. Two factors (no doubt, among others) contribute to this long-term trend of decreasing accidents.

Commanders and other Officers pay close attention to preventable accidents, and work procedures and training methods are under constant scrutiny and adaptation to prevent training and other accidents, on duty and off. That’s one factor, one closely associated with an overall workplace safety ethic, likewise reflected in OSHA and other type safety initiatives. The military, no less than other occupations, has been burdened by an increasing intolerance to any residual risk.

With the same effect but a different motivation, the military has also been constrained by the second factor I’ve seen at work: an almost total reluctance to ever risk military lives unless the risks have been severely mitigated, the benefits are enormous, and the potential cost in lives is infinitesimal. Call it the Powell Doctrine run amok, but it’s been at work in the American Military even before Gulf War One. (In fact, the first Gulf War, Panama, Grenada, Bosnia, and similar military campaigns only reinforced the great American reluctance in all things military.)

Which gets me to the other objections to even having these discussions, that of playing on the turf of the war’s opponents, or somehow giving them a gotcha moment. Look, they pull the data and ignore the facts in giant orders of magnitude more than those who support the war and sometimes get the facts wrong. Our failings won’t give them any more ammunition they wouldn’t find on their own, manufacture, dream up, or distort through their kaleidoscopes of hate.

Any of us in the military know that what we’re really talking about is National Security and the preservation of our way of life. That we fight against multiple enemies, those within and without, who hate America, or reject our ideals, or despise what we represent, or resent our material success, or loath our cultures, or through misguided altruism or idealism, seek to control what God and our founding fathers meant to be free and at Liberty.

Every loss of life in the military piece of that fight represents great sacrifice of treasure, of course felt most deeply by the friends, family, and comrades in arms of those who fall.

But we can’t abandon any battleground in this war against our war. I consider arguing details against those who manipulate and misrepresent, who want to exploit the sacrifices of our military, is a part of a necessary struggle. I don’t think doing so demeans or belittles the military sacrifices. It’s not the only fight. It may not even be a very important one. But it’s a fight MILBLOGS are well suited to wage.

More to come.