Mistaken military death data

I know I make hasty judgments from time to time, but I hope mine are not as egregious as this one. I received the accompany content in an e-mail message today. It purports to show that military deaths were higher during some presidencies than others.

Subject: Fw: Fw: Military death by year and ethnicity

Military Losses, 1980 thru 2007 Whatever your politics, however you lean, however you feel about the current administration, this report should open some eyes.

As tragic as the loss of any member of the US Armed Forces is, consider the following statistics:

The annual fatalities of military members while actively serving in the armed forces from 1980 through 2006:

1980 ………..2,392 (Carter Year)
1981 ………. 2,380 (Reagan Year)
1984 ………. 1,999 (Reagan Year)
1988 ………. 1,819 (Reagan Year)
1989 ………. 1,636 (George H W Year)
1990 ………. 1,508 (George H W Year)
1991 ………. 1,787 (George H W Year )
1992 ………. 1,293 (George H W Year)
1993 ………. 1,21 3 ( Clinton Year )
1994 …… …. 1 ,07 5 ( Clinton Year)
1995 ………. 2,465 ( Clinton Year)
1996 ………. 2,318 ( Clinton Year)
1997 ………… 817 ( Clinton Year)
1998 ………. 2,252 ( Clinton Year)
1999 ……… 1,984 ( Clinton Year)
2000 ………1,983 ( Clinton Year)
2001 ………… 890(George W Year)
2002 ……… 1,007 (George W Year)
2003 ……… 1,410 (George W Year)
2004 ……….1,887 (George W Year)
; ; 2005 …………. 919 (George W Year)
2006………….. 920 (George W Year)
2007……………899 (George W Year)

Clinton years (1993-2000): 14,000 deaths

George W years (2001-2006): 7,932 deaths

If you are surprised when you look at these figures, so was I.
These figures mean that the loss from the two latest conflicts in the Middle East are LESS than the loss of military personnel during Bill Clinton’s presidency; when America wasn’t even involved in a war!

And, I was even more shocked when I read that in 1980, during the reign of President (Nobel Peace Prize winner) Jimmy Carter, there were 2,392 US military fatalities! I think that these figures indicate that many members of our Media and our Politicians will pick and choose the information on which they report. Of course we all know that they present only those ‘facts which support their agenda-driven reporting. But why do so many of them march in lock-step to twist the truth? Where do so many of them get their marching-orders for their agenda?

The data made me curious. Fortunately, later in the document, the author cited the source of the data, so I got it. I learned that there are problems with the data presented here. The data from the original source are not the same as these.

First, and especially problemsome: The data in the e-mail are flat wrong. Table 5 from the Congressional Research Service (CRS) report that the message writer cited shows something different (that report cites its source as “Defense Manpower Data Center, Statistical Information Analysis Division” by the way). Total deaths in 1995, 96, 97, 98, 99, and 2000 were 1,040, 974, 817, 827, 796, and 758, respectively, according to the CRS. All six of those years are reported as substantially higher in the e-mail message. Is it odd that the mistakes all go one direction?

But, there are other problems, too. For example, “military deaths” includes deaths from many causes among people in the military. Deaths are caused by accidents, homicides, terrorist actions, etc. One might wonder whether there are differences across time in the causes of death in the military. No surprise: There are. So, using the actual data, I plotted the percentage of military deaths that the U.S. Department of Defense reported were caused by hostile action between 1980 and (November) 2007. The accompanying images shows that percentage with the base being the full-time equivalent force.

Second, the number of people in the armed services changes over time, too. So a given number of deaths when the military is relatively small would be of greater importance than an equal number of deaths when the military force is much larger. So, a more fitting statistic might be the percentage of deaths based on the size of the force. That’s why I used it as the denominator for the image.

I guess the e-mail message is simply another example of untrustworthy content on the Internet. Sigh.