Category Methodologies

Forecasting U.S. Casualties in Bosnia

Photo by Ssgt. Lisa Zunzanyika-Carpenter 1st Combat Camera Charleston AFB SC
Photo by Ssgt. Lisa Zunzanyika-Carpenter 1st Combat Camera Charleston AFB SC

In previous posts, I highlighted a call for more prediction and accountability in the field of security studies, and detailed Trevor N. Dupuy’s forecasts for the 1990-1991 Gulf War. Today, I will look at The Dupuy Institute’s 1995 estimate of potential casualties in Operation JOINT ENDEAVOR, the U.S. contribution to the North Atlantic Treaty Organization (NATO) peacekeeping effort in Bosnia and Herzegovina.

On 1 November 1995, the leaders of the Serbia, Croatia, and Bosnia, rump states left from the breakup of Yugoslavia, along with representatives from the United States, European Union, and Russia, convened in Dayton, Ohio to negotiate an end to a three-year civil war. The conference resulted from Operation DELIBERATE FORCE, a 21-day air campaign conducted by NATO in August and September against Bosnian Serb forces in Bosnia.

A key component of the negotiation involved deployment of a NATO-led Implementation Force (IFOR) to replace United Nations troops charged with keeping the peace between the warring factions. U.S. European Command (USEUCOM) and NATO had been evaluating potential military involvement in the former Yugoslavia since 1992, and U.S. Army planners started operational planning for a ground mission in Bosnia in August 1995. The Joint Chiefs of Staff alerted USEUCOM for a possible deployment to Bosnia on 2 November.[1]

Up to that point, U.S. President Bill Clinton had been reluctant to commit U.S. ground forces to the conflict and had not yet agreed to do so as part of the Dayton negotiations. As part of the planning process, Joint Staff planners contacted the Deputy Undersecretary of the Army for Operations Research for assistance in developing an estimate of potential U.S. casualties in a peacekeeping operation. The planners were told that no methodology existed for forecasting losses in such non-combat contingency operations.[2]

On the same day the Dayton negotiation began, the Joint Chiefs contracted The Dupuy Institute to use its historical expertise on combat casualties to produce an estimate within three weeks for likely losses in a commitment of 20,000 U.S. troops to a 12-month peacekeeping mission in Bosnia. Under the overall direction of Nicholas Krawciw (Major General, USA, ret.), then President of The Dupuy Institute, a two-track analytical effort began.

One line of effort analyzed the different phases of the mission and compiled list of potential lethal threats for each, including non-hostile accidents. Losses were forecasted using The Dupuy Institute’s combat model, the Tactical Numerical Deterministic Model (TNDM), and estimates of lethality and frequency for specific events. This analysis yielded a probabilistic range for possible casualties.

The second line of inquiry looked at data on 144 historical cases of counterinsurgency and peacekeeping operations compiled for a 1985 study by The Dupuy Institute’s predecessor, the Historical Evaluation and Research Organization (HERO), and other sources. Analysis of 90 of these cases, including all 38 United Nations peacekeeping operation to that date, yielded sufficient data to establish baseline estimates for casualties related to force size and duration.

Coincidentally and fortuitously, both lines of effort produced estimates that overlapped, reinforcing confidence in their validity. The Dupuy Institute delivered its forecast to the Joint Chiefs of Staff within two weeks. It estimated possible U.S. casualties for two scenarios, one a minimal deployment intended to limit risk, and the other for an extended year-long mission.

For the first scenario, The Dupuy Institute estimated 11 to 29 likely U.S. fatalities with a pessimistic potential for 17 to 42 fatalities. There was also the real possibility for a high-casualty event, such as a transport plane crash. For the 12-month deployment, The Dupuy Institute forecasted a 50% chance that U.S. killed from all causes would be below 17 (12 combat deaths and 5 non-combat fatalities) and a 90% chance that total U.S. fatalities would be below 25.

Chairman of the Joint Chiefs of Staff General John Shalikashvili carried The Dupuy Institute’s casualty estimate with him during the meeting in which President Clinton decided to commit U.S. forces to the peacekeeping mission. The participants at Dayton reached agreement on 17 November and an accord was signed on 14 December. Operation JOINT ENDEAVOR began on 2 December with 20,000 U.S. and 60,000 NATO troops moving into Bosnia to keep the peace. NATO’s commitment in Bosnia lasted until 2004 when European Union forces assumed responsibility for the mission.

There were six U.S. casualties from all causes and no combat deaths during JOINT ENDEAVOR.

NOTES

[1] Details of U.S. military involvement in Bosnia peacekeeping can be found in Robert F. Baumann, George W. Gawrych, Walter E. Kretchik, Armed Peacekeepers in Bosnia (Combat Fort Leavenworth, KS: Studies Institute Press, 2004); R. Cody Phillips, Bosnia-Herzegovina: The U.S. Army’s Role in Peace Enforcement Operations, 1995-2004 (Washington, D.C.: U.S. Army Center for Military History, 2005); Harold E. Raugh,. Jr., ed., Operation JOINT ENDEAVOR: V Corps in Bosnia-Herzogovina, 1995-1996: An Oral History (Fort Leavenworth, KS. : Combat Studies Institute Press, 2010).

[2] The Dupuy Instiute’s Bosnia casualty estimate is detailed in Christopher A. Lawrence, America’s Modern Wars: Understanding Iraq, Afghanistan and Vietnam (Philadelphia, PA: Casemate, 2015); and Christopher A. Lawrence, “How Military Historians Are Using Quantitative Analysis — And You Can Too,” History News Network, 15 March 2015.

Estimating Combat Casualties II

Just a few comments on this article:

  1. One notes the claim of 30,000 killed for the 1991 Gulf War. This was typical of some of the discussion at the time. As we know, the real figure was much, much lower.
  2. Note that Jack Anderson is quoting some “3-to-1 Rule.” We are not big fans of “3-to-1 Rules.” Trevor Dupuy does briefly refute it.
  3. Trevor Dupuy does end the discussion by mentioning “combat power ratios.” This is not quite the same as “force ratios.”

Anyhow, interesting blast from the past, although some of this discussion we were also conducting a little over a week ago at a presentation we provided.

 

Estimating Combat Casualties I

Shawn Woodford was recently browsing in a used bookstore in Annapolis. He came across a copy of Genius for War. Tucked in the front cover was this clipping from the Washington Post. It is undated, but makes reference to a Jack Anderson article from 1 November, presumably 1990. So it must have been published sometime shortly thereafter.

19901100EstimatingCombatCasualties

 

Assessing the 1990-1991 Gulf War Forecasts

WargamesA number of forecasts of potential U.S. casualties in a war to evict Iraqi forces from Kuwait appeared in the media in the autumn of 1990. The question of the human costs became a political issue for the administration of George H. W. Bush and influenced strategic and military decision-making.

Almost immediately following President Bush’s decision to commit U.S. forces to the Middle East in August 1990, speculation appeared in the media about what a war between Iraq and a U.S.-led international coalition might entail. In early September, U.S. News & World Report reported “that the U.S. Joint Chiefs of Staff and the National Security Council estimated that the United States would lose between 20,000 and 30,000 dead and wounded soldiers in a Gulf war.” The Bush administration declined official comment on these figures at the time, but the media indicated that they were derived from Defense Department computer models used to wargame possible conflict scenarios.[1] The numbers shocked the American public and became unofficial benchmarks in subsequent public discussion and debate.

A Defense Department wargame exploring U.S. options in Iraq had taken place on 25 August, the results of which allegedly led to “major changes” in military planning.[2] Although linking the wargame and the reported casualty estimate is circumstantial, the cited figures were very much in line with other contemporary U.S. military casualty estimates. A U.S. Army Personnel Command [PERSCOM] document that informed U.S. Central Command [USCENTCOM] troop replacement planning, likely based on pre-crisis plans for the defense of Saudi Arabia against possible Iraqi invasion, anticipated “about 40,000” total losses.[3]

These early estimates were very likely to have been based on a concept plan involving a frontal attack on Iraqi forces in Kuwait using a single U.S. Army corps and a U.S. Marine Expeditionary Force. In part due to concern about potential casualties from this course of action, the Bush administration approved USCENTCOM commander General Norman Schwarzkopf’s preferred concept for a flanking offensive using two U.S. Army corps and additional Marine forces.[4] Despite major reinforcements and a more imaginative battle plan, USCENTCOM medical personnel reportedly briefed Defense Secretary Dick Cheney and Joint Chiefs Chairman Colin Powell in December 1990 that they were anticipating 20,000 casualties, including 7,000 killed in action.[5] Even as late as mid-February 1991, PERSCOM was forecasting 20,000 U.S. casualties in the first five days of combat.[6]

The reported U.S. government casualty estimates prompted various public analysts to offer their own public forecasts. One anonymous “retired general” was quoted as saying “Everyone wants to have the number…Everyone wants to be able to say ‘he’s right or he’s wrong, or this is the way it will go, or this is the way it won’t go, or better yet, the senator or the higher-ranking official is wrong because so-and-so says that the number is this and such.’”[7]

Trevor Dupuy’s forecast was among the first to be cited by the media[8], and he presented it before a hearing of the Senate Armed Services Committee in December.

Other prominent public estimates were offered by political scientists Barry Posen and John J. Mearshimer, and military analyst Joshua Epstein. In November, Posen projected that the Coalition would initiate an air offensive that would quickly gain air superiority, followed by a frontal ground attack lasting approximately 20 days incurring 4,000 (with 1,000 dead) to 10,000 (worst case) casualties. He used the historical casualty rates experienced by Allied forces in Normandy in 1944 and the Israelis in 1967 and 1973 as a rough baseline for his prediction.[9]

Epstein’s prediction in December was similar to Posen’s. Coalition forces would begin with a campaign to obtain control of the air, followed by a ground attack that would succeed within 15-21 days, incurring between 3,000 and 16,000 U.S. casualties, with 1,049-4,136 killed. Like Dupuy, Epstein derived his forecast from a combat model, the Adaptive Dynamic Model.[10]

On the eve of the beginning of the air campaign in January 1991, Mearshimer estimated that Coalition forces would defeat the Iraqis in a week or less and that U.S. forces would suffer fewer than 1,000 killed in combat. Mearshimer’s forecast was based on a qualitative analysis of Coalition and Iraqi forces as opposed to a quantitative one. Although like everyone else he failed to foresee the extended air campaign and believed that successful air/land breakthrough battles in the heart of the Iraqi defenses would minimize casualties, he did fairly evaluate the disparity in quality between Coalition and Iraqi combat forces.[11]

In the aftermath of the rapid defeat of Iraqi forces in Kuwait, the media duly noted the singular accuracy of Mearshimer’s prediction.[12] The relatively disappointing performance of the quantitative models, especially the ones used by the Defense Department, punctuated debates within the U.S. military operations research community over the state of combat modeling. Dubbed “the base of sand problem” by RAND analysts Paul Davis and Donald Blumenthal, serious questions were raised about the accuracy and validity of the methodologies and constructs that underpinned the models.[13] Twenty-five years later, many of these questions remain unaddressed. Some of these will be explored in future posts.

NOTES

[1] “Potential War Casualties Put at 100,000; Gulf crisis: Fewer U.S. troops would be killed or wounded than Iraq soldiers, military experts predict,” Reuters, 5 September 1990; Benjamin Weiser, “Computer Simulations Attempting to Predict the Price of Victory,” Washington Post, 20 January 1991

[2] Brian Shellum, A Chronology of Defense Intelligence in the Gulf War: A Research Aid for Analysts (Washington, D.C.: DIA History Office, 1997), p. 20

[3] John Brinkerhoff and Theodore Silva, The United States Army Reserve in Operation Desert Storm: Personnel Services Support (Alexandria, VA: ANDRULIS Research Corporation, 1995), p. 9, cited in Brian L. Hollandsworth, “Personnel Replacement Operations during Operations Desert Storm and Desert Shield” Master’s Thesis (Ft. Leavenworth, KS: U.S. Army Command and General Staff College, 2015), p. 15

[4] Richard M. Swain, “Lucky War”: Third Army in Desert Storm (Ft. Leavenworth, KS: U.S. Army Command and General Staff College Press, 1994)

[5] Bob Woodward, The Commanders (New York: Simon and Schuster, 1991)

[6] Swain, “Lucky War”, p. 205

[7] Weiser, “Computer Simulations Attempting to Predict the Price of Victory”

[8] “Potential War Casualties Put at 100,000,” Reuters

[9] Barry R. Posen, “Political Objectives and Military Options in the Persian Gulf,” Defense and Arms Control Studies Working Paper, Cambridge, MA: Massachusetts Institute of Technology, November 1990)

[10] Joshua M. Epstein, “War with Iraq: What Price Victory?” Briefing Paper, Brookings Institution, December 1990, cited in Michael O’Hanlon, “Estimating Casualties in a War to Overthrow Saddam,” Orbis, Winter 2003; Weiser, “Computer Simulations Attempting to Predict the Price of Victory”

[11] John. J. Mearshimer, “A War the U.S. Can Win—Decisively,” Chicago Tribune, 15 January 1991

[12] Mike Royko, “Most Experts Really Blew It This Time,” Chicago Tribune, 28 February 1991

[13] Paul K. Davis and Donald Blumenthal, “The Base of Sand Problem: A White Paper on the State of Military Combat Modeling” (Santa Monica, CA: RAND, 1991)

Human Factors in Warfare: Quantifying Morale

Radio broadcast, 1943. Copyprint. Bob Hope Collection, Motion Picture, Broadcasting and Recorded Sound Division, Library of Congress (150)

My colleague Chris Lawrence has taken me to task with regard to my last post on the topic of morale. I asserted that Trevor Dupuy’s claim that human factors influenced the outcome of combat was controversial. Chris pointed out that it was, in fact, Col Dupuy’s claims that 1) human factors in combat could be measured and quantified; and 2) that quantified human factors should be incorporated into combat models, that were really the sources of controversy.

Chris is indisputably correct about that. In my defense, I chose the wording I used in my post deliberately to weasel around an apparently problematic contradiction. While Col Dupuy did indeed believe that human factors in combat could be quantified, he was dubious about the prospect of individually quantifying some of those specific factors, including morale. Since my post was primarily about Dr. Fennell’s work on morale, I side-stepped that issue. Well, today I am going to wade on in.

Col Dupuy first addressed the quantification of human factors in combat in his book Numbers, Predictions & War, published in 1979. His original combat model, the Quantified Judgment Model (QJM), incorporated 73 different variable elements and parameters, including 11 he defined as intangible. By intangible, he meant variables “which are – at least for the present – impossible to quantify with confidence, either because they are essentially qualitative in nature, or because for some other reason they currently defy precise delineation or measurement.”[1] He also believed that “some (such as logistics) may lend themselves to assessment indirectly through the measurement of their effects.”[2]

He divided those intangible variables into three categories[3]:

Sometimes calculable

Probably calculable; not yet calculated

Intangible; probably individually incalculable
  • Combat effectiveness
  • Logistics
  • Leadership
  • Training/experience
  • Initiative
  • Morale
  • Time
  • Space
  • Momentum
  • Intelligence
  • Technology

With regard to leadership, training, and morale, Col Dupuy asserted that

These subjective qualities are almost impossible to assess in absolute terms with complete objectivity. However, the relative capabilities of the opposing leaders in terms of skill, nerve, and determination can probably have more influence on the outcome of a battle than any of the other qualitative variables of combat— if there is a substantial difference in the qualities of leadership of the opposing sides. The same is true, probably to a somewhat lesser extent, if there are substantial differences in the state of training or of combat experience of the two sides, and if there are great differences in their respective states of morale. Accordingly, where solid historical information warrants, these three variables can be given mathematical weights, either individually, or in relationship with the other elements of combat effectiveness, on the basis of professional military judgment, but (under the present “state of the art”) this weighting process is bound to be highly subjective.[4]

Col Dupuy included morale as an independent variable in the QJM’s combat power formula and offered a table of suggested values [5]. However, he did not explain in any detail how to assess morale or apply it in the QJM.

QJM-TNDM Morale Table

As I quoted in my last post, as of 1987, Col Dupuy continued to contend ambivalently that “even though it may not be easily defined and can probably never be quantified, troop morale is very real and can be very important as a contributor to victory or defeat.” He never resolved this seeming contradiction in his writings, but as with several of the intangible variables he identified, he did acknowledge the potential for defining and quantifying some of them. Dr. Fennell’s demonstration of a strong correlation between morale level and rates of sickness, battle exhaustion, desertion, absence without leave and self-inflicted wounds suggests, at least in the British Second Army in northwest Europe in 1944-45, a potential methodology for quantifying morale which could allow its impact to be measured indirectly in much the same manner Col Dupuy measured combat effectiveness. Whether or not the notion of doing so remains controversial will be interesting to see.

NOTES

[1.] Trevor N. Dupuy, Numbers, Predictions & War: Using History To Evaluate Combat Factors And Predict The Outcome Of Battles (New York: The Bobbs-Merrill Company, Inc., 1979), p. 36

[2] ibid., p. 37.

[3] ibid., p. 33.

[4] ibid., p. 37-38.

[5] ibid., p. 231.

Human Factors in Warfare: Measuring Morale

Figure One: Second Army, Weekly Admissions per 1,000 to General Hospitals and Casualty Clearing Stations for Sickness, Battle Exhaustion and SIW, 11 June 1944 to 5 May 1945. Morale scale equivalents are presented on the right hand Y-axis. (Dr. Jonathan Fennell)
Figure One: Second Army, Weekly Admissions per 1,000 to General Hospitals and Casualty Clearing Stations for Sickness, Battle Exhaustion and SIW, 11 June 1944 to 5 May 1945. Morale scale equivalents are presented on the right hand Y-axis. (Dr. Jonathan Fennell)

One of Trevor Dupuy’s more important and controversial contributions to a theory of combat was the assertion that outcomes were dictated in part by behavioral factors, i.e. the human element. Among the influential human factors he identified were leadership, training, experience, and manpower quality. He also recognized the importance of morale.

Morale is an ephemeral quality of military forces and is certainly intangible. Yet even though it may not be easily defined and can probably never be quantified, troop morale is very real and can be very important as a contributor to victory or defeat. The significance of morale is probably inversely proportional to the quality of troops. A well-trained, well-led, cohesive force of veterans will fight well and effectively even if morale is low… Yet for ordinary armies, poor morale can contribute to defeat.[1]

Dr. Jonathan Fennell of the Defence Studies Department at King’s College London recently set out to determine if there were ways of measuring morale by looking at the combat experiences of the British Army in World War II. Fennell proposed

that the concept of morale has no place in a critical analysis of the past unless it is clearly differentiated from definitions associated solely or primarily with mood or cohesion and the group. Instead, for morale to have explanatory value, particularly in a combat environment, a functional conceptualisation is proposed, which, while not excluding the role of mood or group cohesion, focuses its meaning and relevance on motivation and the willingness to act in a manner required by an authority or institution.

Fennell constructed a multi-dimensional model of morale

By drawing on studies made across the social sciences and on primary archival evidence from the British and Commonwealth Army’s experiences in North Africa in the Second World War… It suggests that morale can best be understood as emerging from the subtle interdependencies and interrelationships of the many factors known to affect military means.

Fennell tested his methodology by developing a weekly morale score using bi-weekly censorship summaries of letters and correspondence from members of the British Second Army in the Northwest Europe Campaign in 1944-45.

These summaries proved a useful source to describe and ‘quantify’ levels of morale (through the use of a numerical morale scale). Where morale was described as ‘excellent’, it was awarded a score of 3. ‘High’ morale was given a score of 2 and ‘good’ morale was scored 1. ‘Satisfactory’ morale was given a score of 0 (neither positive or negative). Morale described as ‘severely tried’ was scored -1, while ‘low’ and ‘very low’ morale were scored -2 and -3 respectively.

He then correlated these scores with weekly statistics compiled by the British Second Army and 21st Army Group on rates of sickness, battle exhaustion, desertion, absence without leave (AWOL) and self-inflicted wounds (SIW).

The results of the correlation analysis showed that the tabulated rates (the combined rate of sickness, battle exhaustion, desertion, AWOL and SIW) had an extremely strong negative correlation with morale (-0.949, P<0.001), i.e. when morale was high, sickness rates etc. were low, and when morale was low, sickness rates etc. were high. This is a remarkably strong relationship and shows that these factors when taken together can be used as a quantitative method to assess levels of morale, at the very least for the Army and campaign under discussion.

The results are shown on the graph above. According to Fennell,

This analysis of morale supports the conclusions of much of the recent historiography on the British Army in Northwest Europe; morale was a necessary component of combat effectiveness (morale in Second Army was broadly speaking high throughout the victorious campaign); however, morale was not a sufficient explanation for Second Army’s successes and failures on the battlefield. For example, morale would appear to have been at its highest before and during Operation ‘Market Garden’. But ‘Market Garden’ was a failure. It is likely, as John Buckley has argued, that ‘Market Garden’ was a conceptual failure rather than a morale one. Morale would also appear to have been mostly high during operations in the Low Countries and Germany, but these operations were beset with setbacks and delays.

Fennell further explored the relationship between morale and combat performance, and combat performance and strategy, in his contribution to Anthony King, ed., Frontline: Combat and Cohesion in the Twenty-First Century (Oxford: Oxford University Press, 2015). An earlier version of his chapter can be found here.

NOTES

[1] Trevor N. Dupuy, Understanding Defeat: How To Recover From Loss In Battle To Gain Victory In War (New York: Paragon House, 1990), p. 67

Assessing the TNDA 1990-91 Gulf War Forecast

Map of ground operations of Operation Desert Storm starting invasion February 24-28th 1991. Shows allied and Iraqi forces. Special arrows indicate the American 101st Airborne division moved by air and where the French 6st light division and American 3rd Armored Cavalry Regiment provided security. Image created by Jeff Dahl and reposted under the terms of the GNU Free Documentation License, Version 1.2.
Map of ground operations of Operation Desert Storm starting invasion February 24-28th 1991. Shows allied and Iraqi forces. Special arrows indicate the American 101st Airborne division moved by air and where the French 6st light division and American 3rd Armored Cavalry Regiment provided security. Image created by Jeff Dahl and reposted under the terms of the GNU Free Documentation License, Version 1.2.

[NOTE: This post has been edited to more accurately characterize Trevor Dupuy’s public comments on TNDA’s estimates.]

Operation DESERT STORM began on 17 January 1991 with an extended aerial campaign that lasted 38 days. Ground combat operations were initiated on 24 February and concluded after four days and four hours, with U.S. and Coalition forces having routed the Iraqi Army in Kuwait and in position to annihilate surviving elements rapidly retreating northward. According to official accounting, U.S. forces suffered 148 killed in action and 467 wounded in action, for a total of 614 combat casualties. An additional 235 were killed in non-hostile circumstances.[1]

In retrospect, TNDA’s casualty forecasts turned out to be high, with the actual number of casualties falling below the lowest +/- 50% range of estimates. Forecasts, of course, are sensitive to the initial assumptions they are based upon. In public comments made after the air campaign had started but before the ground phase began, Trevor Dupuy forthrightly stated that TNDA’s estimates were likely to be too high.[2]

In a post-mortem on the forecast in March 1991, Dupuy identified three factors which TNDA’s estimates miscalculated:

  • an underestimation of the effects of the air campaign on Iraqi ground forces;
  • the apparent surprise of Iraqi forces; and
  • an underestimation of the combat effectiveness superiority of U.S. and Coalition forces.[3]

There were also factors that influenced the outcome that TNDA could not have known beforehand. Its estimates were based on an Iraqi Army force of 480,000, a figure derived from open source reports available at the time. However, the U.S. Air Force’s 1993 Gulf War Air Power Survey, using intelligence collected from U.S. government sources, calculated that there were only 336,000 Iraqi Army troops in and near Kuwait in January 1991 (out of a nominal 540,000) due to unit undermanning and troops on leave. The extended air campaign led a further 25-30% to desert and inflicted about 10% casualties, leaving only 200,000-220,000 depleted and demoralized Iraqi troops to face the U.S. and Coalition ground offensive.[4].

TNDA also underestimated the number of U.S. and Coalition ground troops, crediting them with a total of 435,000, when the actual number was approximately 540,000.[5] Instead of the Iraqi Army slightly outnumbering its opponents in Kuwait as TNDA approximated (480,000 to 435,000), U.S. and Coalition forces probably possessed a manpower advantage approaching 2 to 1 or more at the outset of the ground campaign.

There were some aspects of TNDA’s estimate that were remarkably accurate. Although no one foresaw the 38-day air campaign or the four-day ground battle, TNDA did come quite close to anticipating the overall duration of 42 days.

DESERT STORM as planned and executed also bore a striking resemblance to TNDA’s recommended course of action. The opening air campaign, followed by the “left hook” into the western desert by armored and airmobile forces, coupled with holding attacks and penetration of the Iraqi lines on the Kuwaiti-Saudi border were much like a combination of TNDA’s “Colorado Springs,” “Leavenworth,” and “Siege” scenarios. The only substantive difference was the absence of border raids and the use of U.S. airborne/airmobile forces to extend the depth of the “left hook” rather than seal off Kuwait from Iraq. The extended air campaign served much the same intent as TNDA’s “Siege” concept. TNDA even anticipated the potential benefit of the unprecedented effectiveness of the DESERT STORM aerial attack.

How effective “Colorado Springs” will be in damaging and destroying the military effectiveness of the Iraqi ground forces is debatable….On the other hand, the circumstances of this operation are different from past efforts of air forces to “go it alone.” The terrain and vegetation (or lack thereof) favor air attacks to an exceptional degree. And the air forces will be operating with weapons of hitherto unsuspected accuracy and effectiveness against fortified targets. Given these new circumstances, and considering recent historical examples in the 1967 and 1973 Arab-Israeli Wars, the possibility that airpower alone can cause such devastation, destruction, and demoralization as to destroy totally the effectiveness of the Iraqi ground forces cannot be ignored. [6]

In actuality, the U.S. Central Command air planners specifically targeted Saddam’s government in the belief that air power alone might force regime change, which would lead the Iraqi Army to withdraw from Kuwait. Another objective of the air campaign was to reduce the effectiveness of the Iraqi Army by 50% before initiating the ground offensive.[7]

Dupuy and his TNDA colleagues did anticipate that a combination of extended siege-like assault on Iraqi forces in Kuwait could enable the execution of a quick ground attack coup de grace with minimized losses.

The potential of success for such an operation, in the wake of both air and ground efforts made to reduce the Iraqi capacity for offensive along the lines of either Operation “Leavenworth’…or the more elaborate and somewhat riskier “RazzleDazzle”…would produce significant results within a short time. In such a case, losses for these follow-on ground operations would almost certainly be lower than if they had been launched shortly after the war began.[8]

Unfortunately, TNDA did not hazard a casualty estimate for a potential “Colorado Springs/ Siege/Leavenworth/RazzleDazzle” combination scenario, a forecast for which might very well have come closer to the actual outcome.

Dupuy took quite a risk in making such a prominently public forecast, opening his theories and methodology to criticism and judgement. In my next post, I will examine how it stacked up with other predictions and estimates made at the time.

NOTES

[1] Nese F. DeBruyne and Anne Leland, “American War and Military Operations Casualties: Lists and Statistics,” (Washington, D.C.: Congressional Research Service, 2 January 2015), pp. 3, 11

[2] Christopher A. Lawrence, America’s Modern Wars: Understanding Iraq, Afghanistan and Vietnam (Philadelphia, PA: Casemate, 2015) p. 52

[3] Trevor N. Dupuy, “Report on Pre-War Forecasting For Information and Comment: Accuracy of Pre-Kuwait War Forecasts by T.N. Dupuy and HERO-TNDA,” 18 March, 1991. This was published in the April 1991 edition of the online wargaming “fanzine” Simulations Online. The post-mortem also included a revised TNDM casualty calculation for U.S. forces in the ground war phase, using the revised assumptions, of 70 killed and 417 wounded, for a total of 496 casualties. The details used in this revised calculation were not provided in the post-mortem report, so its veracity cannot be currently assessed.

[4] Thomas A. Keaney and Eliot A. Cohen, Gulf War Airpower Survey Summary Report (Washington, D.C.: U.S. Department of the Air Force, 1993), pp. 7, 9-10, 107

[5] Keaney and Cohen, Gulf War Airpower Survey Summary Report, p. 7

[6] Trevor N. Dupuy, Curt Johnson, David L. Bongard, Arnold C. Dupuy, How To Defeat Saddam Hussein: Scenarios and Strategies for the Gulf War (New York: Warner Books, 1991), p. 58

[7] Gulf War Airpower Survey, Vol. I: Planning and Command and Control, Pt. 1 (Washington, D.C.: U.S. Department of the Air Force, 1993), pp. 157, 162-165

[8] Dupuy, et al, How To Defeat Saddam Hussein, p. 114

Forecasting the 1990-1991 Gulf War

DoD photo by Regina Ali
DoD photo by Regina Ali

In my last post on the subject of prediction in security studies, I mentioned that TDI has a public forecasting track record. The first of these, and possibly the most well know, involves the 1990-1991 Gulf War.

On 13 December 1990, Trevor N. Dupuy, President of Trevor N. Dupuy & Associates (TNDA), testified before the House Armed Services Committee on the topic of the looming military confrontation between the military forces of the United States and United Nations Coalition allies and those of Iraq.[1] He offered TNDA’s assessment of the potential character of the forthcoming conflict, as well as estimates of the likely casualties that both sides would suffer. Dupuy published a refined and expanded version of TNDA’s analysis in January 1991.[2]

Based on a methodology derived from Dupuy’s combat models and synthesized data on historical personnel and material combat attrition, TNDA forecast a successful U.S. and Coalition air/ground offensive campaign into Kuwait.[3] Using publicly available sources, TNDA calculated that Iraqi forces in Iraq numbered 480,000, U.S. forces at 310,000, and Coalition allies at 125,000.

The estimated number of casualties varied based on a campaign anticipated to last from 10 to 40 days depending on five projected alternate operational scenarios:

  • Operation “Colorado Springs.” A 10-day air campaign aimed at achieving air superiority and attacking Iraq’s ground forces and war-making infrastructure. While TNDA believed an air campaign would proceed any ground offensive option, Dupuy suggested that it could potentially force an Iraqi surrender without the need for a land attack.
  • Operation “Bulldozer.” A frontal assault on Iraqi forces in Kuwait, lasting 10-20 days.
  • Operation “Leavenworth.” A double envelopment of Iraqi forces in Kuwait using an armored turning force in the west and a U.S. Marine amphibious landing in the east.
  • Operation “RazzleDazzle.” Similar to “Leavenworth,” but combined with an assault along the Iraq-Kuwait border by airborne/airmobile forces for a triple envelopment to encircle all Iraqi forces in Kuwait.
  • Operation “Siege.” A combination of an extended Operation “Colorado Springs” and ground force raids on all of Iraq’s borders. After 10-20 days, one of the three ground attack options (“Bulldozer,” “Leavenworth,” or “RazzleDazzle”) would be initiated to conclude the war.

Based on these assumptions, TNDA produced a range of casualty predictions for U.S. forces that TNDA asserted would probably be accurate to within +/- 50%. These ranged from a low of 380 for a 10-day “Colorado Springs” air-only campaign, to a top-end calculation of 16,645 for a 10-day “Colorado Springs” followed by a 20-day “Bulldozer” frontal assault.

TNDA’s Projection of Likely U.S. Casualties

Scenario Duration

Killed

Wounded

Total

+/-50%
Colorado Springs

10-40 days

190-315

190-315

380-630

—
Bulldozer*

10-20 days

1,858-2,068

8,332-9,222

10,190-11,290

5,335-16,645
Leavenworth*

10-20 days

1,454-1,664

6,309-7,199

7,763-8,863

4,122-12,995
RazzleDazzle*

10-20 days

1,319-1,529

5,534-6,524

6,853-8,053

3,717-11,790
Siege*

10-30 days

564-1,339

1,858-5,470

2,422-6,809

1,451-10,479

* Figures include air casualties

Based on these calculations, TNDA recommended the following course of action:

If the above figures are close to accurate (and history tells us they should should be), then the proper solution is to begin the war with the air campaign of Operation “Colorado Springs.” If this should result in an Iraqi surrender, so much the better. If not, then after about ten days of “Colorado Springs,“ to continue the air campaign for about ten more days while initiating Operation “Siege.” If this does not bring about an Iraqi surrender, the ground campaign should be concluded with Operation “RazzleDazzle.” If this has not brought about an Iraqi surrender, then an advance should be made through the desert to destroy any resisting Iraqi forces and to occupy Baghdad if necessary.[4]

In my next post, I will assess the accuracy of TNDA’s forecast and how it stacked up against others made at the time.

Notes

[1] Armed Services Committee, U.S. House of Representatives, Testimony of Col. T. N. Dupuy, USA, Ret. (Washington D.C.: 13 December 1990)

[2] Trevor N. Dupuy, Curt Johnson, David L. Bongard, Arnold C. Dupuy, If War Comes, How To Defeat Saddam Hussein (McLean, VA.: HERO Books, 1991); subsequently republished as How To Defeat Saddam Hussein: Scenarios and Strategies for the Gulf War (New York: Warner Books, 1991).

[3] These are the Quantified Judgement Model (QJM) and Tactical Numerical Deterministic Model (TNDM). Dupuy’s methodological approach and his first cut on a Gulf War estimate are described in Chapter 7 of Trevor N. Dupuy, Attrition: Forecasting Battle Casualties and Equipment Losses in Modern War (McLean, VA.: HERO Books, 1990).

[4] Dupuy, et al, How To Defeat Saddam Hussein, 126

History News Network (HNN)

hnn-logo-new

We do have a half-dozen links listed down at the bottom of the right hand column of this blog. One is the History News Network

I have five articles posted on HNN. Two of them being posts from this blog. They are:

How Military Historians Are Using Quantitative Analysis

Did the Pentagon Learn from Vietnam?

Defeating an Insurgency by Air

Did I Just Write the Largest History Book Ever?

Are Russians Really Long-Suffering

Now, they do choose the headlines, and sometimes that gives a different feel to the article. So for example, one of my blog posts was titled “Russian Revolutions.” The exact same article on the HNN is titled “Are Russians Really Long-Suffering.” This apparently got a couple of people up in arms because the article did not talk about all the famines and oppression in Russia and the Soviet Union. It did not, because it was about revolutions, and in particular was about revolutions that succeeded. The famines in the 1890s, 1920s and 1930s did not directly lead to a successful revolution (a point that I think is pretty significant).

The article “Did I Just Write…” is actually a shorter version of an article I posted on the Aberdeen Bookstore website: Long version of “Did I Just Write…” Part of the reason that I wrote that article was to see if someone would come out of the woodwork and post that there was a larger book published (usually these postings start with something like “the author is an idiot because….”). I did not get that for this article. This does sort of confirm my suspicion that this is indeed the largest single volume history book ever written (no disrespect intended for the 11-volumes done by the Durants…which were four million words and 10,000 pages). I wonder if this is something I should submit to the Guinness Book of World Records? Will I get free beer for that?

 

Screw Theory! We Need More Prediction in Security Studies!

Johnny Carson as Carnac the Magnificent; taken from the January 24, 2005 broadcast of The Tonight Show.
Johnny Carson as Carnac the Magnificent; taken from the January 24, 2005 broadcast of The Tonight Show.

My previous post touched on the apparent analytical bankruptcy underlying the U.S. government’s approach to counterterrorism policy. While many fingers were pointed at the government for this state of affairs, at least one scholar admitted that “the leading [academic] terrorism research was mostly just political theory and anecdotes” which has “left policy makers to design counterterrorism strategies without the benefit of facts.”

So what can be done about this? Well, Michael D. Ward, a Professor of Political Science at Duke University, has suggested a radical solution: test the theories to see if they can accurately predict real world outcomes. Ward recently published an article in the Journal of Global Security Studies (read it now before it goes behind the paywall) arguing in favor of less theory and more prediction in the fields of international relations and security studies.

[W]e need less theory because most theory is an attempt to rescue or adapt extant theory. We need more predictions in order to keep track of how well we understand the world around us. They will tell us how good our theories are and where we need better explanations.

As Ward explained,

[P]rediction is deeply embedded in the philosophy of science… The argument is that if you can develop models that provide an understanding—without a teleology of why things happen—you should be able to generate predictions that will not only be accurate, but may also be useful in a larger societal context.

Ward argued that “until very recently, most of this thread of work in security studies had been lost, or if not lost, at least abandoned.” The reason for this was the existence of a longstanding epistemological disagreement: “Many social scientists see a sharp distinction between explanation on the one hand and prediction on the other. Indeed, this distinction is often sharp enough that it is argued that doing one of these things cuts you out of doing the other.”

For the most part, Ward asserted, the theorists have won out over the empiricists.

[M]any scholars (but few others) will tell you that we need more theory. Doubtless they are right. Few of them really mean “theory” in the sense that I reserve for the term. Few of them mean “theory” in the sense of analytical narratives. Many of them mean “detailed, plausible stories” about how stuff occurs.

In light of the uncomfortable conclusion that more detailed, plausible stories about how stuff occurs does not actually yield more insight, Ward has adopted a decidedly contrarian stance.

I am here to suggest that less is more. Thus, let me be the first to call for less theory in security studies. We should winnow the many, many such “theories” that occupy the world of security studies.

Instead, we need more predictions.

He went on to detail his argument.

We need these predictions for four reasons. First, we need these predictions to help us make relevant statements about the world around us. We also need these predictions to help us throw out the bad “theories” that continue to flourish. These predictions will help drive our research into new areas, away from moribund approaches that have been followed for many decades. Finally, and perhaps most important, predictions will force us to keep on track.

But making predictions is only part of the process. Tracking them and accounting for their accuracy is the vital corollary to improving both accuracy and theory. As Ward pointed out, “One reason that many hate predictions is that talking heads make many predictions in the media, but few of them ever keep track of how well they are doing.” Most, in fact, are wrong; few are held accountable for it.

Of course, the use of empirical methods to predict the outcomes of future events animated much of Trevor N. Dupuy’s approach to historical analysis and is at the heart of what The Dupuy Institute carries on doing today. Both have made well-documented predictions that have also been remarkably accurate. More about those in the next post.