Military Fatalities in Iraq

Take a look at this state ranking. It ranks states by Iraqi-war casualties per 100,000 residents. The chart was made as part of a rather silly debate about red states and blue states that doesn’t interest me. What interests me is Utah.

Utah is at the very bottom. Why? I don’t know actually. But here are 3 factors to keep in mind:

1. Utah gets a little messed up in everything that is “per capita” when the characteristic in question is only found among adults. We have more children— about a third of the state is under 18, compared to a fourth for the nation. So our denominator has all those kids in it who aren’t in the military.

2. Do we have fewer people serving per capita? I have no idea, but would be surprised if Utahns tended to not join the military. One obvious problem is that military service often happens right when missions do. Is this the explanation? Are we not joining the military because we go on missions instead? That would be very interesting.

3. Perhaps we have as many people serving in the military, but fewer of those have been called over to Iraq. For example, maybe Utah’s military capacity is not geared for Iraqi sorts of conflicts.

Controlling for those above factors, do we still look good? I don’t know but I admit to certain native skepticism. The child thing is only going to move us to about .44, — still a very low casualty rate. If Utahns do join the military in typical numbers, and do go to Iraq in typical numbers, well we’ve been blessed to have few fatalities in Iraq; may that trend continue and extend to all the other soldiers who are in harm’s way.

45 comments for “Military Fatalities in Iraq

  1. Check with someone who actually understands statistics and probability. All of those numbers except from large states like Texas seem pretty small to me and may not be statistically significant, or in other words the variation can be explained by random chance alone.

  2. Another possiblity is random variation. A 95% confidence interval for Utah’s “true” Iraq-death rate goes all the way up to .76, which would put us right in the middle of the chart.

    Still, my guess is that you’re on to something with #2.

    (Nit-pick…the chart refers to deaths, not casualties. Casualties generaly means dead and wounded combined.)

  3. I see that Mike makes the same point. BTW, Mike, Frank is an empirical economist and he definitely understands statistics and probability, (and I’m actually surprised that he didn’t bring it up himself.)

  4. ed,

    Thanks for the correction. What are you basing the confidence interval on? Did you just assume a normal distribution or are you using a poisson that fits the hazard nature of the data?


    Assuming ed calculated the CI right, A 95% confidence interval that dead-ends at the average means that there is some reasonable but small chance that Utah is average (a 2.5% chance for that side on a two-sided test). So you’re right that I should have added that in as another factor and I ‘m glad you mentioned it. The odds that Utah is at the mean would actually be pretty small, but it is certainly one possibility.

  5. Utahns who do join the military tend not to join the combat arms, as far as I can tell, which one is more likely to join when one is 18. After the mission, one is more likely to go for something that requires a little more sense and a little less youthful vigor.

    I’ve also seen that a lot of middle class Utahns support the military in the abstract but look on actual military service as suspect and declasse–the Tommy Atkins syndrome-but I would be very surprised if the rest of the country weren’t the same way.

  6. This is, of course, related to to Ed’s point; We a ranking something from low to high. There is a reasonable chance Utah’s rate is measured with negative error ( in the mu + epsilon sense, not in the incorrect reporting sense; Frank and Ed you know what I mean) and Vermont’s with postive error. Seems like this is an appropriate time to use a Bayesian shrinkage estimator.

  7. Frank: I used a poisson.

    Rereading my comment #2, I notice that I referred to Utah as “us.” That’s strange, because I’ve never lived in Utah except for two and a half years at BYU. I’ve often had other people assume I was from Utah, but I didn’t realize until today that even I assume I’m from Utah.

  8. Karl,

    Lars Lefgren always wants to use a Bayesian shrinkage estimator too. It must be something in the Chicago water.


    Thanks for the info. That makes a lot of sense – and combines 2 and 3 in a reasonable way.

  9. Another big issue is the type of reserve and National Gaurd units based in Utah. Reservists and National Guard troops at one point made up fifty percent of the troops in Iraq. Reservists and National Guard troops have suffered an inordinate number of casualties and members of the same unit are from the same geographical area. So if one National Guard unit from Ohio gets hit, you get a lot of Ohio casualties. So if we look at where the National Guard and reserve units from Utah have served in Iraq, that should explain alot of the casualty figures.

    I think, but am not sure, that most of the NG and reserve units in Utah are intelligence, special forces, and artilery units. I seem to recall that the special forces units served in Afganastan but not Iraq. There is also a Marine reserve unit with troops from Utah and Southern Nevada that served in the invasion but went home before the casualties started to mount. I don’t think that the intelligence units have suffered high casulaty rates. I know that a large reserve artilery unit from So. Utah was just deployed in Iraq, but I don’t know what they are doing and I haven’t heard about any deaths or injuries from that unit.

    If large numbers of Utah reservists and NG troops are in non-combat missions that would explain the low death rates, even if the military participation rates among Utahns is the same as the average for the rest of the country.

  10. Frank,

    Bayesian shrinkage is a beautiful thing, and the Chicago the water is fresh which naturally gives us an advantage over saltier areas.

    Lars, mentioned Bayesian shrinkage? I thought his standard response was, “use a regression discontinuity approach.”

  11. Karl, he’s come a long way. (Not that he doesn’t still love RD)


    Thanks for the info. What you say seems consonant with Adam’s comments, perhaps? It also provides an interesting and plausible explanation that relates to Utah’s mormon-ness. If mormons tend to join the military post-mission and therefore don’t go into combat units, then it is no surprise that Utah’s units would tend to not be combat units.

  12. Frank,

    I disagree with Adam’s comments a little because artilery and special forces are clearly combat units. Returned missionaries with language skills do tend to join a special interpretation and interrogation reserve unit based in SLC, but I think that unit only has about 500 members. I would bet that military participation rates in Utah exceed the national average. I would also bet that Utahns join combat units at high rates. I just think Utahns have been very lucky that thier reserve units have not served in as many dangerous missions as other NG and reserve units.

  13. Let me point out a notable exception to the point about the predominance of non-combat Utah military units: there is a reserve company of Marine infantrymen at Fort Douglas, Salt Lake City, that shrouded itself in glory in the invasion of Iraq in 2003. My brother was among their number, and wrote about it here: They suffered one combat fatality.

  14. Interesting post. This is similar to an analysis that I’ve kept since hearing the Church’s opponents claim that Utah’s suicide rate is higher than the national average and then conclude horrible things about the restored gospel.

    In 2001, the national suicide rate was 10.8 and Utah’s was 14.1 per 100/k residents. Refining the analysis reveals that:
    1. Utah’s average is higher than the nation’s
    2. Most of the states with rates higher than the national average are in the West, inland. Maybe isolation contributes to suicide.
    3. Every state that touches Utah has a higher suicide rate than Utah’s. Another nearby state, MT, also has a higher rate than Utah’s. There appears to be something about Utah that counters the local causes of a high suicide rate.
    4. Of the states that touch Utah, the suicide rates have a negative correlation with the LDS %. The notable outlier is NV, which has has other unique factors that may lead to suicide. It appears that the LDS influence is counter to suicide, not a contributor.

    The numbers:
    National rate: 10.8
    Utah’s rate: 14.1
    States contiguous with Utah: AZ, 14.5; Colo, 16.3; ID, 15.9; NM, 19.8; NV 18.3; WY, 16.8.
    States noncontiguous with Utah that have higher rates than Utah: AK, 16.1; AR, 14.2; MT, 19.3; OK, 14.8; OR, 14.5; WV, 15.9.
    Source:, p. 106

  15. “Returned missionaries with language skills do tend to join a special interpretation and interrogation reserve unit based in SLC, but I think that unit only has about 500 members.”

    Actually, the 300th MI Brigade might be bigger than that, I think (I belonged to it but I’m not an expert). Anyway, its a substantial portion of the Army National Guard in Utah. The Army puts so many MI units in Utah because so many returned missionaries sign up who already speak various foreign languages.

  16. I am also unclear on how deaths are allocated to states. For example, if I’m in the Utah National Guard but claim my residence in some other state, or vice versa, to which state is my death assigned? What about the men on active duty? Is it their state of enlistment, the state they claim as residency, or what?

  17. A lot of people who join the armed forces get assigned to a military base in a different state. So the question of how the state association with a combat death is made is an important one. For example, air force casualties have been far lower than marine casualties, and as far as I know, Utah has a lot of air force and very few marines.

    I would assume that the state association is not based on the state of residence at enlistment, but rather the state of residence when deployed, which is the same as asking, “In what state is the military base located where the deceased unit is stationed?”

    And that, of course, is an entirely meaningless evaluation, since the location of bases is largely based on congressional pork-seeking and the needs of the military. It has little to do with the local population’s support of the military.

  18. I can confirm that many of the Utah reservists were artillery which was oriented towards the Korean situation. They were all sent to guard the chemical weapons depot although they got relieved a year or two ago there. I had a few brothers-in-law who were stuck out at the base there. Which wasn’t as bad as Iraq for sure. They were often worried about being sent out. But it never happened.

  19. Adam Greenwood:

    “After the mission, one is more likely to go for something that requires a little more sense and a little less youthful vigor.”

    I think I understand the point you’re making, but this statement is not very respectful to the 18 and 19 year olds currently dying in Iraq. They didn’t have the sense not to enlist in the infantry? So, maybe I’m not quite sure I understand what you meant by this.

    God bless Cindy Sheehan and others who suffer tragic losses every day as a result of this conflict. My thoughts and prayers are with you all.

  20. There is a national guard unit from my area, I’m not sure of the numbers, but hundreds of guys from our town and surrounding areas are now in Iraq. Some fathers and sons, brothers, it’s scary.

    This same unit was sent to Korea and saw combat, but not one was killed. There’s a video about it called, I think, “They All Came Back.” My uncle was there.

    I’ve said this elsewhere, but we have hopes for the same result with Iraq. I think it was the presence of the priesthood in such numbers.

  21. Frank, I’ll admit I was lukewarm when you originally came on as permanent blogger (for little good reason), but you have become the first and only bona fide social scientist on the blog–Russell may have it in him but has never really done social scientific posts. Kudos. If we can now just get you to do more on sociology of religion…

  22. gst, thanks for the link!

    manaen, there are a long list of similar sorts of claims that can be explained by carefully looking at the data. Thanks for doing the legwork on that one.

    Anne, I hope it turns out well.


    ” I’ll admit I was lukewarm when you originally came on as permanent blogger (for little good reason)”

    I bet I can guess

    ” If we can now just get you to do more on sociology of religion…”

    Will the economics of religion do? And I have to agree that as far as the social science of mormonism goes, the field is white already to harvest.

  23. I agree, red vs. blue is silly, but — and I’m just posing a what if here — what if the statistics were solid, completely defensible?

  24. I think we’ve overlooked something important here about the social scientific study of religion and modern warfare.

    What’s really interesting about that chart is that it suggests that Utahns–which, we all know, is just social science shorthand for “Mormons”–are protected from harm or accident in combat. We already know from the experience of the Nephites with the Gadianton Robbers that the preaching of the word unto one’s enemies can be an effective defensive measure. But I think we’ve all missed the military potential of faith as a force multiplier of offensive capabilities.

    What if in, say, attacking an insurgent stronghold like Fallujah, American forces were to suffer casualties at the rate of our lovely Deseret, instead of at the rate of heathen Vermont? Samuel the Lamanite may have been merely trying to preach to wicked Nephites, but his take-home message for today is all about outfitting soldiers with the ultimate in body armor. Talk about a book written for our times! Sure, the Nephites were prepared with their helmets and breastplates and lo, even their scimitars, but their cost-benefit analysis of the situation didn’t neglect the strategic advantage of spiritual preparation, either.

    Most non-Mormons are going to be skeptical. Clearly, the only thing to do is to organize a double-blind experiment, where half the soldiers in a brigade, chosen at random, are ordered to taken the discussions, gain a testimony, and be baptized before being sent on a combat mission to Iraq. A comparison of Mormon and non-Mormon casualties, as well as a regression analysis of various indicators of faithfulness and obedience to the commandments, should show fairly rapidly the effectiveness of guardian angels against IED’s. Frank, which statistical analysis best corrects for distortion due to bad things happening to good people? Is there a commonly accepted “this life is a test” test?

    Really, there’s no reason that the US Army shouldn’t be God’s Army, too, with a corresponding casualty rate.

  25. You’re being sarcastic, Jonathan Green, but against whom? Most people here seem a little embarrassed by Utah’s low figures, if anything. You shouldn’t squander dudgeon as if it were a renewable resource.

  26. Miss E.,

    There are sensible virtues, you bet, but heroism and bravery are virtues without being sensible. They are virtues most often found in the young.

  27. Adam, I have no target, no dudgeon, no sarcasm. It’s just the kind of thing that strikes me as funny two days before the semester starts.

    Besides, haven’t you heard that we’ve reached the tipping point on humor? From here on out, it’s only going to get harder and harder to get a laugh out of people. With the traditional sources drying up, not to mention how M* has revoked all permits for humor on their territory, it seemed wasteful to leave comedic potential unexploited. They say that the rising cost of humor will spur technological innovation and make it economically attractive to sustain conversation based only on hot air or other gaseous mixtures, but I would like to postpone that point for as long as possible.

  28. I’ve read your post a few times now, and I’ve concluded that it’s not funny. I suppose it’s theoretically possible to write something funny about this subject, but I can’t think of what it might be. You’d be wiser to not even make the attempt.

  29. On one of the military blogs, numbers were posted that indicate the death rate in Iraq is really very low comparably. However, I think this doesn’t reflect the terrible injuries being incurred by our troops. When I saw a brief part of something, I think it was on PBS, showing grievously injured troops, loss of limbs and terrible burns, and paralysis, that just broke my heart.

    As a mother, I’m terribly saddened by this war more than any other in my life. I was a teenager during Vietnam, but not very informed. Now I look at it from a different perspective and it breaks my heart. I don’t think we can pull out, but those poor kids….and I know Iraquis are dying every day. I know they were, but my heart rests ultimately with those young service people from my town–with the possibility my grandsons could be over there in a few years, oh, how do people stand the worry? That is more reality for me than the news stories.

  30. I am no statastician, but there are a lot of various factors that influence casualties. Of course, enlistments and where the untis are from are a big factor, but I am skeptical about the scientific (or non-scientific) assumptions that are made about a “state’s” contribution to a war effort based on the total of casualties. The real question is how many soldiers per capita have served from each state. Even so, I am not sure that one state is better than another because of the amount of casualties.

    Still, if we we really are trying to make inferences about casualties, 1872 casualties spread over 50 states of origin of various sizes and among units in Bahrain, Kuwait, The Gulf of Aden, Oman, The Gulf of Oman, The Arabian Sea, The Red Sea, Iraq, Quatar, and the Persian Gulf (all areas of Operation Iraqi Freedom) which numbers include hostile and non-hostile deaths (like equipment injuries, traffic accidents and sickness) is not very scientific. If there were more casualties (like the number WWII or even like Vietnam, we could add more “dots” on the graph to try and start plotting something.

    However, seeing that at the current rate of casualties in Iraq it will take 73.4 years before we even meet the Vietnam threshold, I think any inferences made at this point are pure conjecture.

  31. Adam is right re: types of units in utah. Also, one should add the air national guard units (low casualties of course) and the army engineering units. These are construction units. Not exactly front-line either.

    re: # of Utahns in Guard & Reserve. Utah actually has a very high percentage per capita serving. We have far more units than average given our population. This is because of effectively political lobbying, lots of open space, and most importantly: Utahns are willing to enlist.

  32. I’d just like to say that until about 3 months ago I was working onsite at the HQ of the Army National Guard. I think Jason may be onto something with the 300th Military Intelligence Brigade(?). I think the type of National Guard units that Utah has may make a difference. Artillery and Military Intelligence units are unlikely to be converted to MP units (or do MP duties) orused for Transportation, both of which are higher risk. States in the west get artillery units more often since they have more room for practice ranges, small states & territories get MP, Medical & Transportation Companies.

    The variety of languages that the church teaches its missionaries at the MTC don’t really include the languages of Iraq (Arabic/Kurdish) or Afghanistan (Pashto & others). What they *would* have in relatively high numbers (is by reasoning, not from classified info) is Spanish & Korean, both of which while being the language of a country that the USA has *issues* with (Cuba & North Korea) they are also the language of a country where the church sends a decent number of missionaries (Mexico/Argentina & South Korea).


  33. Here’s a site that gives total number of troops deployed per state, as well as number of casualties.

    Data may be older here. It says Utah had 4664 deployed and 7 deaths, 1.5 per 1000. Puts Utah pretty low, at 38th or so, in terms of deaths per actively deployed troops.

    Nationwide, the number was 1.8 per 1000.

    From this data, there is certainly no significant evidence that Utahns in Iraq are less likely to die then other soldiers in Iraq.

  34. Yet the fact remains that our national guard unit, which saw significant combat in Korea, did not lose one man. I don’t think it has anything to do with being from Utah, but I do think there were no atheists in those foxholes. There must be some power to believing in and relying on God.

  35. Annegb says: “On one of the military blogs, numbers were posted that indicate the death rate in Iraq is really very low comparably. However, I think this doesn’t reflect the terrible injuries being incurred by our troops. When I saw a brief part of something, I think it was on PBS, showing grievously injured troops, loss of limbs and terrible burns, and paralysis, that just broke my heart.”

    It is true that many American soldiers have experienced very traumatic injuries in the Iraq war. Just so you know, one of the reasons that there are so many amputations and other grievous injuries is because fewer soldiers are dying of their wounds. This, in turn, is explained by two important factors. First, soldiers now have better body armor, which protects their vital organs. Unfortunately, their faces and limbs are not as well protected– although the next generation of defensive armor is designed to correct this. Second, medicine is much better now, meaning that soldiers now are more likely to survive with a lost limb instead of bleeding to death on the battlefield. It is tragic that hundreds of soldiers have come home as amputees, but it is good that they are alive.

    Here are some links where you can read more:

  36. Yeah, I knew that. I was commenting on how tragically the war has affected lives, which I hadn’t realized before.

    I can’t remember if I asked this before, I think it was somewhere else, but what about the draft? Do we need more men there?

    I just had a letter published in the paper calling for the draft, but now I may have changed my mind. Probably made some hit lists.

  37. Frank: Thanks, amigo. Much agape. It is, indeed, there now. Perhaps I was put on the comment watch list after I tried to threadjack one of Nate’s posts a few weeks ago… Or maybe it’s Al Gonzalez and John Poindexter keeping an eye on me. I found out during a recent trip to Utah for a brother’s wedding that I am on the administration’s no-fly list. Or, rather, some guy with the same name as me who’s like 20 years older than I am. Kind of a bummer. I always get the super search treatment.

    Anne: As a PhD student in national security studies, I am not aware of anyone in the field who thinks we need a draft to win this war. (Some scholars have been recommending a return to the draft for the last twenty years for sociological reasons, but that’s a different story.) More men in Iraq would solve some problems but cause others. At some point, the problems caused by more soldiers outweigh the benefits. American military planners believe that increasing the number of boots on the ground would also increase the resentment of Iraqis towards the occupying force. This is part of the reason they are working so hard to train and prepare Iraqi forces to share the burden. Iraqi forces may be less capable but they make Iraqi civilians much, much happier. As Iraqi forces increase and American (and Coalition) forces are pulled out, it will become more and more obvious that the troubles in Iraq are not caused by rapacious foreign military occupiers but by foreign terrorists, criminals, and Baathist holdouts. For this reason, most American military planners want fewer American troops in Iraq (and Afghanistan), not more.

  38. Interesting thread. A couple points to verify what others have commented on. First at the start of Iraqi freedom, Utah lead the nation in percentage of Reserve component troops deployed. We had %88 percent of our National Guard and Army Reserve personnel deployed somewhere. Not all were overseas, but the following groups were, the two battalions of the 300th MI Brigade that are located in Utah were deployed to Iraq (with the exception of one company which was in Bosnia), as were a large portion of the 19th SF group(in Afghanistan), two of our three Engineer units(in Iraq), the 145th field Artillery was deployed to guard the Deseret Chemical Depot and Dugway Proving Grounds. The 211th aviation had aircraft in Afghanistan, and the 222 Field Artillery was mobilized for Iraq but did not get out of the US, as they were re-directed to Ft Lewis.

    The 222nd FA is now in Iraq, and contrary to someones thoughts about what Artillery would be doing, they have been mostly re-trained as MP’s and are kicking down doors. The MI personnel also operate in a very exposed manner(there is a small element of MI there now, attached to other units).

    And we must also remember the Foxtrot Company USMC Reserves. From Utah/Nevada. They fought all the way up to Bagdad. With very few casualties considering the intensity of the fighting they were involved in.

    With all the reserve personnel deployed, most the fatalities of Utahns have been Active duty, only three or four since we went into Afghanistan and Iraq, have been reserve component troops.

    As to the Draft, a draft would cause severe problems. Our military is not structured to handle conscripts. We have an all volunteer force that would have serious difficulties absorbing draftees.

  39. Let us not forget Marine Staff Sergeant James Cawley who was a platoon sergeant with Fox Company, 2nd Battalion, 23rd Marines. He was killed in an accident in Iraq.

  40. In an earlier post, Jonathan Stone stated: I would assume that the state association is not based on the state of residence at enlistment, but rather the state of residence when deployed, which is the same as asking, “In what state is the military base located where the deceased unit is stationed?”

    I am currently on Active Duty and this statement is incorrect (but there is another reason not yet discussed).

    First off, a lot of those deployed are in the National Guard, which must become residents of the state they are serving. They do not get the special residency status, since they are relatively stationary (deployments being the only exception, which do not count as “permanent changes of station”).

    But in relation to Active Duty, there is something known as “Home of Record”. When you first enter active duty, the Home of Record is the state you entered active duty from. You are given the choice to maintain that state as your home of record, or change your home of record to the state you are currently stationed in (it must be a permanent duty assignment, ie PCS, not a temporary duty assignment, ie TDY).

    The decision nearly always boils down to taxes in reality. If your state taxes are less in the state you entered active duty from than they are in the state you are currently stationed in, then you will most likely keep your residency in the state you entered active duty from. But if the state you are currently stationed in has lower taxes, or no income tax at all, then most people will switch over and realize the tax savings.

    There is one other important (unrelated) factor, so I don’t want to make it sound like taxes are the only basis for the decision. But they are an important one relative to this discussion, since Utah has such a high income tax burden on the lower middle class compared to most states, and almost all military members are in the lower middle class income bracket. (Some lower enlisted are in the “poor” bracket, and some high-ranking officers are of course in higher classes, but a vast majority are in the lower middle class.)

    For example, I entered active duty from Idaho, and now I am in Maryland. So I get the choice right now of whether to be an Idaho resident or a Maryland resident. Since Idaho does not tax active duty income if you are stationed outside the state, I have no tax burden in Idaho, and when weighing other factors it is best if I remain an Idaho resident. But I could make Maryland my state of residency–I don’t have to, though.

    If I PCS’d to Florida, then I have the choice of keeping my current residency or switching to Florida. For me, that translates into keeping Idaho residency or switching to Florida residency. I could not decide at that time to become a Maryland resident, since I had no Maryland affiliation anymore.

    If I went TDY anywhere, I cannot change residency, since TDYs are shorter duration.

    So the process is not arbitrary (you cannot just pick your favorite state that you lived in years ago at any time), nor is it necessarily based on current duty station. It comes down to relative tax burden, and likelihood of wanting to move back to that state when leaving the military. Since Utah is high on the tax burden for lower middle class, that partially explains the numbers (but not completely, remember National Guard are different; and there are also to poor Utahns that get sent to Taxachusetts or California for their first duty assignment).

    I am sure missions also have something to do with it, but I just wanted to throw out an additional factor.

  41. John David Payne: I will probably print your post and quote from it when I write the letter saying I’ve changed my mind. Sigh. I also went on an Iraqi blog and they said the same thing. Sigh…

    The 222nd is the group from my town. Many friends there–my son’s best friend, who enlisted after his death (following?) is in Iraq, active duty Marines, training Iraqi policemen.

    I have been trying to find soldier blogs from Iraq, but they can’t blog now, or only to other soldiers or something. I would really like to get their point of view. I understand the government’s decision there, but it’s hard to understand the issue with only the media to interpret it.

Comments are closed.