"It Made a Lot of Sense to Kill Skilled Workers": The Firebombing of Tokyo in March 1945
Author(s): Thomas R. Searle
Source: The Journal of Military History, Vol. 66, No. 1 (Jan., 2002), pp. 103-133
Published by: Society for Military History
Stable URL: http://www.jstor.org/stable/2677346
Accessed: 04/11/2010 00:49
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=smh.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].
Society for Military History is collaborating with JSTOR to digitize, preserve and extend access to The Journal
of Military History.
http://www.jstor.org
"It Made a Lot of Sense to Kill
The Firebombing
SkilledWorkers"1:
ofTokyoin March 1945
Thomas R. Searle
QNthe nightof9-10 March 1945, the UnitedStates ArmyAir Forces
(USAAF) conducted the most destructiveair raid in history.The
targetwas Tokyo and, by the time the firesdied out the next morning,
nearlysixteen square miles of the city were destroyed,leaving at least
83,793 Japanese civiliansdead, more than 40,918 injured,and over one
millionhomeless.2The raid was a turningpointin the bombingofJapan.
Before9 March,most raids were "precision raids" thatused high-explosive bombs against Japanese factories,killingfew Japanese civilians.
After9 March, the USAAF devoted the bulk of its effortto "area raids"
that used incendiarybombs to burn down Japanese cities and to kill
hundredsof thousandsofJapanese civilians.
Historians have not devoted as much attentionto the incendiary
bombingofJapan as theyhave to eitherthe WorldWar II bombingcampaign in Europe or the atomic bombingof Japan,but it has hardlybeen
ignored.3The literaturehas focused,however,on the sequence of oper1. U.S. Air Force Oral HistoryProgram,interviewof Lt. Gen. Ira C. Eaker,22
May 1962, K239.0512-627,AirForce HistoricalResearchAgency,MaxwellAirForce
Base, Alabama(hereafter
citedas HRA).
2. These disconcertingly
precisecasualtyfiguresare theofficialcountmade by
the Tokyopolice duringthewarand are almostcertainlyan underestimation.
news in U.S. newspapers
3. The incendiarybombingof Japan was front-page
duringthe war. Since the war,a steadystreamof books thatare whollyor largely
FrankCravenand
devotedto the bombingofJapanhas appeared,including:WVesley
James Lea Cate, The ArmyAir Forces in World War II, 7 vols. (Chicago: Universityof
Chicago Press, 1953); MartinCaidin,A Torchto theEnemy(New York:Ballantine,
AirForce
1960); WilburH. Morrison,
PointofNo Return:The StoryoftheTwentieth
(New York:TimesBooks,1979); RobertGuillain,I Saw TokyoBurning:An EyewitThzeJournal of Military History 66 (January 2002): 103-34
C Societv forMilitarvHistor
*
103
TIIOMAS R. SEARLE
ations ratherthan on long-classifiedplanningdocuments,and this bias
has led historiansto misunderstandseveral aspects ofthe campaign.Historians also misunderstandthe relationshipbetween bombing tactics
used againstGermanyand those used againstJapan.
This essay willoffera new interpretationof the U.S. strategicbombing campaign againstJapan. I shall argue firstthat the incendiarybombing of Japanese cities was not a radical departure fromthe way the
USAAF attacked Germanyin WorldWar II; second, thatthe shiftto area
bombingwas a continuationof the attack on Japanese industryand not
an abandonment of attacks on industryin favorof attackingJapanese
morale; and third,that the shiftfromprecision attacks on factoriesto
area attacks on major Japanese cities had been part of U.S. plans for
years and was encouraged by some of the same factorsthat led the
ness NarrativefromPearl HarbortoHiroshima,trans.WilliamByron(GardenCity,
N.Y.: Doubleday,1981); HaywoodS. Hansell,Jr.,StrategicAir WarAgainstJapan
(MaxwellAir Force Base, Ala.: AirpowerResearch Institute,1980); Haywood S.
Hansell,Jr.,The StrategicAir WarAgainstGermanyand Japan: A Memoir(Washington:OfficeofAirForceHistory,
1986); KevinHerbert,
MaximumEffort:
TheB-29s
AgainstJapan (Manhattan,Kans.: SunflowerUniversityPress,1983); Hoito Edoin,
The NightTokyoBurned:The IncendiaryCampaign AgainstJapan,March-August,
1945 (New York:St. Martin'sPress,1987); CurtisE. LeMay and Bill Yenne,Superfortress:The Storyof theB-29 and AmericanAir Power (New York:McGraw-Hill,
1988); E. BartlettKerr,Flames Over Tokyo:The U.S. ArmyAir Forces' Incendiary
CampaignAgainstJapan, 1944-1945 (New York:Donald I. Fine, 1991); KennethP.
XVerrell,
BlanketsofFire: U.S. BombersOver Japan During WorldWarII (Washington:SmithsonianInstitution
Press,1996); Daniel L. Haulman,HittingHome: TheAir
OffensiveAgainst Japan (Washington:Air Force Historyand Museums Program,
1999). Thereare also a numberofgeneralworkson strategicbombingin WorldWVar
II thatdevotea significant
partof theirargumentsto the bombingofJapan,including:R. J.Overy,TheAirWar,1939-1945 (Chelsea,Mich.:ScarboroughHouse,1980);
RonaldSchaffer,
AmericanBombingin WorldWarHI(NewYork:
WingsofJudgment:
OxfordUniversity
Press,1985); MichaelS. Sherry,The Rise ofAmericanAirPower:
The CreationofArmageddon(New Haven,Conn.: Yale University
Press,1987); Conrad C. Crane,Bombs, Cities,and Civilians:AmericanAirpowerStrategyin World
WarII (Lawrence:University
PressofKansas,1993); Geoffrey
Perret,WingedVictory:
TheArmyAir Forces in WorldWarII (New York:RandomHouse, 1993); Benjamin
FranklinCooling,ed., Case Studies in tle AchievementofAir Superiority(Washington:GPO, 1994); R. CargillHall,ed.,Case Studiesin StrategicBombardment(NWashofseveralkeyleadersin
ington:GPO, 1998). Finally,biographiesand autobiographies
the bombingofJapaninclude:HenryH. Arnold,Global Mission (New York:Harper
and Brothers,1949); CurtisE. LeMaywithMacKinlayKantor,Mission withLeMay
HAP: The StoryoftheU.S.
(GardenCity,N.Y.:Doubleday,1965); ThomasM. Coffey,
Air Force and the Man Wh-o
Built It, General HenryH. "Hap" Arnold (New York:
Iron Eagle: Th-eTurbulentLifeofGeneral Curtis
Viking,1982); Thomas M. Coffey,
The Quest: HaywoodHansell and
LeMay (NewYork:Crown,1986); CharlesGriffith,
AmericanStrategicBombingin WorldWarII (MaxwellAirForce Base, Ala.: AirUniversityPress,1999); and Dik Alan Daso, Hap Arnoldand theEvolutionofAmerican
SmithsonianInstitution
Airpower(WVashington:
Press,2000).
104
*
THE JOURNAL OF
"ItMade a LotofSensetoKillSkilledWorkers"
USAAF to conductarea incendiarybombingofGermancities. I shall also
show thatthe shiftdid not resultfromchanges ofcommand (Major General Curtis E. LeMay's replacingBrigadierGeneral Haywood S. Hansell).
Finally,I shall demonstratethat Japanese civilian casualties were not
accidental or incidental,but an explicitgoal of the incendiaryraids on
Japanese cities. Since this argumentmustnecessarilybe imbeddedin its
historicalcontext,I shall begin by reviewingthe ideas behind strategic
bombing, the strategicbombing campaign against Germany,and the
bombingcampaign againstJapan beforeMarch 1945.
StrategicBombing of Germany
When the United States enteredthe Second WorldWVar
in December
1941, General Henry H. (Hap) Arnold, Commanding General of the
USAAF,and much of his staffwere committedto a concept of airpower
whichheld thatairplanes could serve as powerful"independent"tools of
war that mighteven win wars single-handedlythrough"strategicbombing,"i.e., air attackson an enemy'sheartlandthatwould destroyhis will
and capabilityto wage war.This would be a new typeofwarfare,one that
the U.S. Army(to which the USAAF then belonged) and the U.S. Navy
were unable to conduct. IfUSAAF strategicbombingcould make a major
contributionto the war effort,
it would providea powerfulargumentfor
the service autonomy U.S. airmen had sought fordecades. Thus, dramatic results fromstrategicbombing became the means by which the
USAAF hoped not only to help win the war,but also to pursue its postwar goals.4
The USAAF also had a tactical doctrineforhow to conduct strategic
bombing:large formationsof heavily armed bombers would fighttheir
way throughthe enemy'sdefensesand conduct veryprecise bombingin
daylightto destroycarefullyselected targets,the loss of which would
This doctrinewas
cripple the enemy'swar effortquickly and efficiently.
based on the idea that industrialeconomies were delicate webs of interdependentfactories,powerplants,and transportationlinks. The USAAF
believed that the loss of a small number of critical plants could bring
down an entire industryand the loss of a small number of industries
4. By the timethe UnitedStatesenteredthe Second WNorld
WVar,
everyonerealized thatairplaneswereimportantmilitaryweapons.Whatwas stillopen to discussion was how they could best be used. Arnoldand the other strategicbombing
advocatesbelievedthattheyhad the answerto thatquestion,and manyoutsidethe
USAAFagreed.TheirargumentswereconvincingenoughthatPresidentFranklinD.
Roosevelt,SecretaryofWNar
HenryL. Stimson,and ArmyChiefofStaffGeorgeC. Marshall-among otherswithnothingto gain froman independentair force-werewillingto let the USAAFconductlarge-scalestrategicbombingcampaigns.
MILITARY HISTORY
*
105
TIIOMAS R. SEARLE
would cripple war productionand even cause a general economic collapse. When the United States entered the war,the USAAF believed its
B-17 and B-24 bombers could fighttheir way to the heartlands of the
German Reich and the Japanese Empire. In the Norden bombsightthe
USAAF feltthat it had the aiming system necessary to hit and destroy
the criticalelementsofthe Germanand Japanese war economies. Before
the war, however,the USAAF was not sure it would have bases close
enough to Japan and GermanyforB-17s and B-24s, so the USAAF began
a programto develop a bomberwitha much longerrange: the B-29.5
By the timethe UnitedStates recoveredfromthe shock oftheJapanese attack on Pearl Harbor and could consider offensiveoperations,
Japan had conquered so manyPacificIslands and such a largepartofthe
Asian mainland thatit would be years beforethe Japanese home islands
were within range of U.S. bombers. Even withoutthe Allies' strategic
decision to defeatGermanyfirst,Germanywas necessarilythe firsttarget of the USAAF because it was withinrange of Allied bases in Britain
(fromwhich the Britishhad been bombingGerman-occupiedareas for
some time). Britishdoctrine and planes, however,were different,
and
particularlyafter Air Chief Marshal ArthurT. Harris took charge of
BritishBomber Command in February1942, the Britishpursued a policy of "area" bombing German cities at night.6An area bombing raid
attemptedto destroylarge partsof an industrialcity in orderto weaken
the enemy by destroyingthe factories,governmentoffices,and workers'
homes withinthe targetarea. Whereas precision bombingattemptedto
destroya small numberof criticalplants and industries,area bombing,
in contrast,attemptedto damage all industries.
In theory,precisionbombingappears to be much more efficientthan
area bombing,but operational problems made precision bombing less
effectivethan its advocates expected. Findingthe critical point in the
enemy'swar industryrequiredextremelydetailed and accurate information about the nation'seconomy,informationhe was tryinghard to keep
secret. When the targetswere in factcritical,the enemy defendedthem
fiercely,made them harder to hit by dispersingthem or puttingthem
underground,and rebuiltthemquicklywhen theywere hit. Area bombing,on the other hand, did not require such accurate intelligenceabout
5. On the originsoftheB-29,see Cravenand Cate,ArmyAirForces, 5: 3-32.
6. The decisionto emphasizedestruction
ofGermancitiesdid notoriginatewith
Harris,but he executed the policywithsuch vrigor,
tenacity,and success thatit is
widelyassociatedwithhim.For a detailedcomparativestudyofUSAAFand British
bombingdoctrinebefore1945, see Tami Davis Biddle,"Rhetoricand Realityin Air
Warfare:The Evolutionof Britishand AmericanIdeas about StrategicBombing,
1917-1945" (Ph.D. diss.,Yale University,
1995); and also her "Britishand American
Approachesto StrategicBombing,"Journalof StrategicStudies 18 (March 1995):
91-144.
106
*
THE JOURNAL OF
"ItMade a LotofSensetoKillSkilledWorkers"
the enemy,and since area targetswere much largerthan precision targets,area bombingcould be done effectively
bycrewsand equipmentthat
could not hit precision targets.Area bombingcould also be done under
conditionssuch as nightand bad weatherthatmade the targetsharderto
hitbut also made the bombersharderforthe defendersto findand attack.
The main operationaldrawbackofarea bombingwas thata huge number
ofbombs had to be droppedbeforethe generallevel ofdestructionmade
an appreciable impact on the enemy's war effort.Since the area targets
were almost always cities, area bombingentailed significantcasualties
among the civilianpopulation.Such casualties raised ethical issues that
precisionbombingmightavoid,and when casualties were "friendly"civilians (for example, Frenchmen killed by Allied bombing of occupied
France), theyposed immediatepoliticalproblemsas well.7
By early 1943, the United States was finallyassembling enough
planes and crews in Britainto make a serious contributionto the strategic bombingof Germanybut had not yet achieved much. BritishPrime
MinisterWinstonChurchillwas convinced that the Britishwere on the
righttrack with area bombing. He wanted the USAAF to abandon its
plans for a daylightprecision bombing campaign separate from the
Britishnightarea campaign and instead join and reinforcethe British
campaign. Churchillsaid as much to PresidentFranklinD. Roosevelt at
the Casablanca Conferencein January1943, and Roosevelt tentatively
agreed to this change in tactics. This agreement,in the words of the
USAAF's officialhistory,created a "crisis."8The change would forcethe
USAAF to renounce its plans and doctrine,retrainits crews,and modify
its aircraft;forexample, flamescame out of the exhaust pipes of the B17s, making them easy to find in the nightsky. From a bureaucratic
standpoint,the change was unacceptable because the USAAF,which was
tryingto escape the controlof the U.S. Army,would then fallunder the
controlof the Royal Air Force. General Arnoldchose not to oppose the
Presidentor the PrimeMinisterhimselfand instead sent the commander
of USAAF bomberforcesin Britain,Major General Ira Eaker,to attempt
to change the Prime Minister'smind. Eaker succeeded.9 One USAAF
7. For example,the Allies wereveryconcernedabout Frenchcasualties from
Alliedbombing,and PrimeMinisterWinstonChurchilltriedto get PresidentRooseveltto changeGeneralDwightD. Eisenhower'splans forthebombingin supportof
the D-day landings.For a relateddiscussion,see RichardG. Davis, "GermanRail
Yards and Cities: U.S. BombingPolicy,1944-1945," Air Power History,Summer
1995, 53-55.
8. Cravenand Cate,ArmyAirForces, 2: 115.
9. BritishsourcesindicatethatChurchill'sstaffhad alreadyconvincedhimthat
the USAAF had to be allowedto attemptdaylightbombing,but thatsimplymeans
thatEaker and the USAAFhad convincedthemand theyconvincedChurchill.For
related discussion,see Charles Websterand Noble Frankland,The StrategicAir
MILITARY HISTORY
*
107
TIIOMAS R. SEARLE
planner went so faras to claim that this was "the crucial turningpoint
in the conduct of the war in Europe."10
Having overcome the Britishthreat,the USAAF now had to defeat
the German air force.It turnedout that the B-17s and B-24s could not
fighttheir way throughGerman defenses without sufferingcrushing
losses, and in late 1943, the U.S. bombers had to stop bombingtargets
deep inside Germanyuntil long-rangefighterswere available to escort
them.11With the help of fighterescorts, the USAAF was finallyable to
defeatthe German air forcein early 1944. But weatherstillcaused problems. The Norden bombsightenabled the USAAF to bomb withimpressive accuracy in clear weather,but the weatherover Germanywas rarely
clear. Radar offereda solution.The Britishhad developed a radar system
called H2S that enabled bombers to look throughthe clouds and distinguish between water,land, and built-upareas, but not much else. The
RadiationLaboratoryat the MassachusettsInstituteofTechnologydeveloped an improvedversion called H2X; a few were available foruse in
USAAF bombers by the fallof 1943, but these could not providea clear
enough picture of the groundto allow the bombardierto hit a specific
factory.In the words of the officialhistorians,this was "a type of area
bombing"-a type in which only about 4 percent of the bombs landed
within a mile of the aiming point.12 With such poor accuracy, the
bombers needed targetsthat were several miles wide. Effectively,
that
meant aimingat the centersof cities ratherthan at individualfactories.
In the last eighteenmonthsof the war in Europe, the USAAF launched
at least sixty-ninesubstantialraids (each comprisedof at least one hundred heavy bombers), which dropped a total of almost sixty thousand
tons ofbombs on targetsdesignatedas the "cityarea" oftwenty-five
differentGerman cities.3 The USAAF actually did a great deal more area
bombingof German cities than these statisticssuggest,under the guise
of attackingGerman transportation.As the officialhistoriansacknowledged,withsuch inaccuratebombing,"the aimingpointbecame a highly
theoreticalterm."1'4
Those with "theoretical,"ethical, or other reasons
fornot wantingto targetGerman city areas could aim at German rail
Offensive
AgainstGermany,1939-1945 (London:HMSO, 1961), 1: 360-63; and Biddie, "Rhetoricand Reality,"533-37. For his part, the PrimeMinisterencouraged
Eaker to believethathis presentation
had won the USAAFits reprieveand tookthis
tack in his memoirs.WinstonChurchill,The Second WorldWar,6 vols. (New York:
HoughtonMifflin,
1950), 4: 678-80.
10. Hansell,Germanyand Japan, 69-7 1.
11. Cravenand Cate,ArmyAirForces, 2: 705.
12. Ibid.,3: 20.
13. Davis, "GermanRail Yardsand Cities,"48.
14. Cravenand Cate,ArmyAirForces,3: 20.
108
*
THE JOURNAL OF
"ItMadea LotofSensetoKillSkilledWVorkers"
yards since these were typicallylocated in the center of German cities.
Althoughthe USAAF also conducted many radar-aimedattacks against
German rail yards, everyone involved understoodthat there was little
differencebetweencityarea raids and radar-guidedraids on rail yardsin
termsof planning,execution,or results.USAAF commandersessentially
acknowledgedthis factby using a large percentageof incendiarybombs
(the preferredweapon against cities) on these raids even though such
bombs were ineffective
against rail yards,the officialtargets.15
This did not representan abandonment of precision bombingas a
goal. Instead, the USAAF accepted that under certain weather conditions, it could not hit small precision targets.When cloud cover over
Germanymade precisionbombingimpossible (nearlyhalfthe time) the
USAAF conducted area bombingratherthan no bombing.Much of the
literatureon U.S. strategicbombingin WorldWar II is devoted to trying
to explain why the USAAF conducted area raids againstJapan but did
not conduct themagainstGermany.Since the USAAF in factdid a great
deal of area bombing in Germany,these works need to be revised or
abandoned entirely.16
The B-17 and B-24 were impressive aircraftand the Combined
Bomber Offensiveagainst Germany was dramaticallymore impressive
than any previousbombingcampaign. But the Allies did not defeatGermany with airpower alone; the Soviet Red Army,the Allied forces in
Italy,and the D-day invasionofFrance,forexample,provedgroundcombat to be farfromout ofdate. Since Arnoldand the advocates ofan independent U.S. Air Force could never be sure just what theyhad to do to
gain independence, they feltenormous pressure to do ever more with
strategicbombing.This pressurewas keenlyfeltin the B-29 programand
the strategicbombingofJapan.
Operations against Japan
War,the design,development,testing,and
Duringthe Second WVorld
production of the B-29 cost over three billion dollars-half again as
15. Davis, "GermanRail Yardsand Cities,"passim. Of course rail yardswere
themdid disrupttheGermanwar
important
targetsin theirownright,and destroying
economy,butas earlyas 1941,one ofthereasonstheBritishwereattackingrailyards
was that the bombs thatmissed the targetwould stillland in Germancities. For
relateddiscussion,see Th-eStrategicAir WarAgainstGermany1939-1945,Reportof
Husby MichaelBeethamand JohnWV.
thze
BritishBombingSurveyUnit,forewords
materialby Sebastian Cox (1946; reprint,London: Frank Cass,
ton, introductory
1998), 5.
16. This line of debate shouldhave died when the officialhistoriansacknowledged USAAF area bombingin 1949, but somehowit lived on into the 1990s. For
example,see Crane,Bombs,Cities,and Civilians, passim.
MILITARY HISTORY
*
109
TIIOMAS R. SEARLE
much as the two billion dollars spent on the atomic bomb-and made
substantialdemands on wartimeresources and manpower.'7 The USAAF
gamble" because the B-29
referredto this as "the three-billion-dollar
Arnoldand the USAAF
wentinto mass productionbeforeits firstflight.'8
took an enormous risk in devotingsuch vast resources to a plane that
had not been tested,but Arnoldfeltthatthe B-29 mightachieve the sort
of spectacular successes that would advance his notion of strategicairpower. As he wrote to one of his subordinates, "The B-29 project is
importantto me because I am convinced that it is vital to the futureof
the ArmyAir Forces."19RushingB-29 developmentalmost guaranteed
that the early planes would experience enormous technical problems,
but it gave the USAAF the B-29 beforethe war ended.
By late 1943 the USAAF had decided to use the B-29 onlyagainstthe
Japanese because its long rangewas essential in the Pacificbut less critical in Europe. Arnold was determinedthat he and the USAAF would
decide how to use these bombers with as littleinterferenceas possible
fromthe Armyand Navy.To achieve this he created a unique organization called "TwentiethAir Force" to conduct B-29 operations.The TwentiethAirForce was directlysubordinateto theJointChiefsofStaff(JCS).
General Arnoldacted as the executive agent forthe JCS and personally
commanded the TwentiethAir Force fromWVashington,
D.C. Though the
B-29s would be based in areas under the control of the theater commandersand would receive logisticalsupportfromthe theaters,Twentieth Air Force was independent of all of the theater commanders and
controlledall B-29 combat operations.20As a memberof the JCS,Arnold
could turndown requests forsupportfromthe Allied theatercommanders (Admiral Louis Mountbatten,General Joseph W. Stilwell,Admiral
ChesterW. Nimitz,and General Douglas MacArthur)-keepingthe B-29s
on theirindependent,strategictargets-while demandingsupportfrom
the theaters.21One of the keys to selling this unusual command struc17. For the impactof the B-29 programon wartimemanpowerconcerns,see
Centerof Military
ed., CommandDecisions (Washington:
Kent RobertsGreenfield,
History,1990), 375.
18. Cravenand Cate,ArmyAir Forces, 5: 7. The USAAFhedgedits bet on the
signinga contractwithConsolidatedto producethe
BoeingB-29 by simultaneously
As it turnedout, the B-32 had even worsetrouB-32 to meet similarspecifications.
to theAlliedwareffort.
bles thanthe B-29 and made almostno contribution
19. Arnoldto LeMay,9 December1944, letter,OfficialDecimalFiles370.2, 20th
AF,Box 106, p. 1, HenryH. ArnoldPapers,ManuscriptDivision,LibraryofCongress,
D.C.
Washington,
20. The main subordinateunits under the TwentiethAir Force were the XX
BomberCommand,operatingout ofChinaand India,and theXXI BomberCommand,
operatingout oftheMarianas.
theater,servedas chiefofstaff
21. Stilwellcommandedthe China-Burma-India
to ChiangKai-shek,and also heldthepositionofdeputycommanderoftheSouthEast
110
*
THE JOURNAL OF
"ItMade a LotofSensetoKillSkilledWorkers"
ture to the JCS was the argumentthat the long range of the B-29 meant
that it would flyfrombases in several theatersagainst targetsin Japan,
which was outside all of the theaters,and thatunityof command in the
base areas (under the theatercommanders) should be sacrificedin favor
of unityof command over the targetarea (under Arnold).22
Beginningin mid-1944,the USAAF based the firstB-29s in India and
attackedJapan fromstagingbases in China. Later thatyear largerforces
began operatingfrombases builton the newlycapturedMariana Islands
(Guam, Saipan, and Tinian). The enormouslogisticaldifficulties
ofoperatingfromChina (into whichfueland bombs had to be flownfromIndia
over the Himalayas) were apparent even beforethe firstB-29s arrived.
Worse yet, the most importanttargetsin Japan could not be reached
fromthe parts of China then under Allied control.As a result,the Air
Staffsaw the Marianas as the crucial base forB-29 operations against
Japaneven beforethe TwentiethAir Force was created. In fact,the main
reasons forconductingoperations fromChina were to make up fornot
having provided Chiang Kai-shek with more supportearlier in the war
and because those were the onlyAllied bases withinrangeof the Japanese home islands at the time the firstB-29 units were ready.23
Operations againstJapan fromChina began on 14 June 1944 under
the command of BrigadierGeneral KennethB. Wolfe,and fromthe Marianas on 24 November 1944 under the command of General Hansell.
Wolfewas one of the USAAF's finestproductionand engineeringofficers
and had personallysupervisedthe B-29 program.Hansell had played an
importantrole in planningthe bombingof both Germanyand Japan and
had commanded a B-17 wingin Europe forsix months.He was chief of
staffof the TwentiethAir Force beforetakingcommand in the Marianas.
Operations fromChina did not go well, so ArnoldfiredWolfeon 4 July
1944 and the next month replaced him with Major General Curtis E.
LeMay.24But even witha new commander,no one expected greatresults
fromChina. Disappointmentsthere were balanced by the hope that
operationsfromthe Marianas would turnthingsaround,but earlyresults
fromthe Marianas were no better.
Nimitzand MacArthur
Asia Command,headed by Mountbatten.
headed the Central
Pacificand SouthwestPacifictheaters,respectively.
22. Hansell,Germanyand Japan, 157-58.
23. Cravenand Cate,ArmyAirForces, 5: 25, 29, 30. These operationsalso provided an importantsource of combat experienceforuse in futureoperationsand
plans. Werrell,BlanketsofFire, claims thattheirmain contribution
to the war was
thattheygave GeneralLeMayexperiencerunningB-29 operations.
24. Wolfewas promotedand returnedto the productionfield.LeMay did not
arriveuntil29 Augustbecause he insistedon learningto flytheB-29 beforeshipping
out to India. Forrelateddiscussion,see Hansell,StrategicAir WarAgainstJapan, 45,
and LeMayand Kantor,MissionwithLeMay,323, 324.
MILITARY
HISTORY
*
111
THOMAS R. SEARLE
The B-29s were attemptingto destroyhigh-priority
industrialtargets
in Japanwithhigh-altitude,
daylightprecisionbombing,but by mid-January 1945, theyhad not destroyeda singletargetin Japan.The principal
cause of bombinginaccuracy was the weatherover Japan,in particular
high-altitudewinds of over two hundred knots and persistentheavy
cloud cover. Beforethe TwentiethAir Force startedhigh-altitudeflights
over Japan,the USAAF was unaware of the jet streamthere and was in
no way prepared forthe problems it posed to bombingaccuracy. The
Norden bombsightcould not compensate forcrosswindsof such magnitude. Bombingrunsconducted downwindhad groundspeeds ofover five
hundredmiles per hour,makingit impossibleforbombardiersto line up
theirsightsin time. Flyinginto the wind devoured fueland leftthe aircraft over the target,exposed to antiaircraftfire,for too long.25 The
extremewinds could cause even more serious problems.As one participant recalled, "[T]he damn targetbacked rightoffthe radar; we were
going backward over ground."26Heavy cloud cover meant that crews
were rarelyable to do visual bombing.As in Europe, radar was the only
way ofbombingthroughthe clouds, but the radar thenused by the B-29
(the AN/APQ-13,withperformancecomparable to the H2X used against
Germany)did not providea clear enough pictureto be effectivein precision bombing.27
By Januaryof 1945, even the TwentiethAir Force Headquarters in
Washingtonfeltthat "the TwentiethAir Force [and the B-29] has been
That monthArnold
ineffectiveas an instrumentof strategicwarfare."28
respondedby firingthe commanderof the B-29 operationsin the Marianas, General Hansell. Arnoldreplaced him withGeneral LeMay as part
of his decision to stop B-29 operations fromChina and India and consolidate them in the Marianas. Nevertheless,Arnoldcontinuedto worry
that "unless somethingdrasticis done to change [the poor results]soon
it will not be long before the B-29 is just another tactical airplane,"
25. RichardH. Kohn and JosephP. Harahan,StrategicAir Warfare:An Interview with Generals CurtisE. LeMay,Leon W Johnson,David A. Burchinal,and
JackJ. Catton(WVashington:
OfficeofAirForce History,1988), 53-55.
26. USAF Oral HistoryProgram,interviewwithGeneralDavidA. Burchinal,11
April1975, K239.0512-837,p. 60, HRA.
27. Hansell,Germanyand Japan, 231. Startingin late June 1945, the 315th
Bombardment
radarthatwas
Wingarrivedfromthe Stateswiththe new AN/APQ-7
able to produceprecisionresults(a highpercentageofbombswithinthreethousand
feetofthe target)againsttargetsthatwerechosen because theywereeasilyidentifiable on radar.
28. Lt. Col. R. S. McNamara(StatisticalControlOffice,20th AF) to Col. C. E.
Combs (DeputyChiefof Staff,20th AF) 16 January1945, memorandum,Subject:
of the TwentiethAir Force as a StrategicWVeapon,
Effectiveness
AF File #760.310A
(20th AF,OperationsAnalysis),1944-45, Air Force HistorySupportOffice,Bolling
Air Force Base WVashin.6tonn
D.C.
112
*
THE JOURNAL OF
_"It
Made a Lot ofSense to Kill Skilled Workers"
divertedfromstrategicbombingto supportingArmyand Navy forces.9
Arnold'sbiographersuggeststhatdistressover the B-29 campaign caused
Arnold'smassive heartattack in January1945.30 By the end of February
resultswere stillpoor and the B-29 looked more and more like an expensive and embarrassingfailure.
LeMay understoodthat in order to keep his job, he would have to
finda way to succeed. The obvious solutionwas to switchfromprecision
bombing to area bombing, which, given the known flammabilityof
Japanese cities, would mean using incendiaries ratherthan high explosives. In late February,with the support of TwentiethAir Force Headquarters in Washington,LeMay tried area incendiarybombing. On 25
February1945, he launched the largestB-29 raid to date against Tokyo,
droppingover fourhundred tons of incendiarybombs on the city. The
raid burned about one square mile of Tokyo. It was the most destructive
raid against Japan to date, but it did not produce results capable of
changingthe directionofa campaign.One moredecision had to be made
beforethe USAAF was ready to mountthe raid that resultedin the 9-10
March conflagration.
Duringthe 25 Februaryraid,the B-29s flewas theywere designedto
fly-i.e., in daylight,in formation,at high altitude. Their altitude protected them fromantiaircraftfire,and the crossfiresproduced by their
machine guns when flyingin formationprotected them fromenemy
fighters.The major technologicalinnovationsof the B-29 (remote controlmachine gun turretsand pressurizedcrew areas) were designedwith
these tactics in mind.
For the 9 March raid, LeMay abandoned accepted tactics, stripped
the B-29s of theirmachine guns, and sent them over Tokyo at low altitude in a single-filestreamratherthan in a formation.The only concession he made to Japanese defenseswas to flyat nightwhen most of the
were ineffective.
LeMay concealed his change in tactics
Japanese fighters
fromhis superiors,perhaps (as he says) to protectthem frompossible
failure,but perhaps because he also feared they mightveto his radical
new idea.31
By changing tactics, LeMay risked decimating his force, but he
solved all of his otherproblems.At low altitudetherewas no jet stream,
and the targetswere much easier to hit with inaccurate incendiary
bombs. "Hitting"an area targetwith incendiarybombs meant getting
enough of the incendiaries to land close enough togetherso that the
numerous small fires merged into one vast, uncontrollable fire. The
29. Arnoldto Brig.Gen. Lauris Norstad(Chiefof Staff,20th AF), 14 January
1945, letter,citedin Coffey,
HAP,357-58.
30. Coffey,
HAP,338, 358.
31. LeMayand Kantor,Mission WithLeMay,348.
MILITARY HISTORY
*
113
THOMAS R. SEARLE
weightthat the B-29s saved by not carryingtheir guns, gunners,and
ammunitiontranslated into an additional three thousand pounds of
incendiarybombs per aircraft,contributingapproximately25 percentto
the totalamount ofbombs dropped on 9 March. Flyingat low altitudein
singlefileused less fuelthan climbingto high altitudeand flyingin formation,permittingan additional three thousand pounds of incendiary
bombs per aircraft.32
Thus, his new tactics enabled LeMay to double the
bombload of each plane and also to ensure that the bombs would land
closer together.By flyingat night,he mightjust get away withflyinglow
withoutmachine guns. LeMay naturallyfearedthat the Japanese might
devise countermeasuresin response to his change in tactics,so ifthe 9
March raid worked,he planned to take fulladvantageof the initialsurprise by running four more "maximum effort"incendiary raids as
quicklyas possible.33
This raid also marked a radical change in the results of USAAF
bombingofJapan. Previously,raids were typicallyflownat highaltitude
and dropped high explosives that failed to hit factoriesand killed relativelyfew Japanese civilians. From 9 March on, the bulk of the B-29
effortwent into low-altitudeincendiaryraids that burned down Japan's
cities and killed hundredsof thousands of Japanese civilians.A change
in tactics and targetselectionhas rarelyproduced such a radical change
in results.The operationalchronologiesof the B-29 campaign and some
postwar memoirs have, however,led historians to draw several erroneous conclusions. One is that the primarygoal of USAAF incendiary
bombingof Japan (like Britishincendiarybombingof Germany)was to
"demoralize the urban population."34In fact,Japanese industrywas the
primarytargetof the area raids, as it was forprecision raids. Another
mistakenview is that the leaders of the USAAF did not fullyaccept the
fact that they were killinglarge numbers of civilians. Some observers
claim that the leaders somehow concealed fromthemselvesthe factsof
what theywere doing; others maintainthat the civilian casualties were
unintended"collateral damage."35But in fact,the leaders of the USAAF
knew exactly what theywere doing,and civilian casualties were one of
the explicitobjectives of area incendiarybombingapprovedby both the
USAAF and the JointChiefsofStaff.The B-29 crews also understoodthis
32. Cravenand Cate,ArmyAirForces, 5: 612-13.
33. TacticalMissionReport,Box B26, CurtisE. LeMayPapers,ManuscriptDivision,LibraryofCongress;MissionNumber40, flown10 March1945.
34. Overy,Air War,1939-1945, 99.
35. Forexample,Sherry,TheRise ofAmericanAirPower,leans towardthe"self
delusion"school. The officialhistoriansof the ArmyAir Forces,Cravenet al., lean
towardthe "collateraldamage"viewofJapaneseciviliancasualties.
114
*
THE JOURNAL OF
"ItMade a LotofSensetoKillSkilledWVorkers"
because theirordersforthe 9 March raid explicitlylistedJapanese civilian casualties as one of the goals of the raid.36
A thirderroris to exaggerateLeMay's role in choosingto attack area
targets.3The attacks on Japanese area targetsbegan in earnest shortly
afterLeMay took over command in the Marianas. The success of the 9
March raid (and othersbased on that pattern)made LeMay's career,but
he did not discover these targetson his own out on Guam in March of
1945. These area targetswere included in USAAF plans in 1943, long
beforethe B-29s startedbombingJapan, and the intentionwas to start
destroyingthem in the middle of the bombing campaign (ideally in
March 1945). A variationon this misunderstandingis to claim that General Arnolddecided to switchto area bombingin late 1944 and decided
to replace Hansell with LeMay because Hansell would not conduct area
bombingand LeMay would. In fact,Hansell had already conducted area
bombing of Japanese cities, and if the B-29s had remained under
Hansell's command, they would still have bombed urban area targets.
Further,Arnoldhad otherreasons forreplacingHansell withLeMay and
would have done it soon, irrespectiveof theirviews on area bombing.A
carefulstudy of USAAF plans forbombingJapan and the details of the
campaign conducted againstJapan will bear this out.
USAAF Plans to AttackJapanese Cities and Civilians
It would be idle to search forthe USAAF planner who firstthought
of making incendiaryattacks on Japan. The vulnerabilityof Japanese
cities to fireswas well known to anyone with even a casual interestin
Japan and had been amply illustratedby the Tokyo earthquake and fire
of 1923. As Americanspondered the possibilityofwar withJapan,many
considered taking advantage of this vulnerability.In 1932, retired
BrigadierGeneral William (Billy) Mitchell,one of the nation's foremost
advocates of airpowerand a virtualfolkhero to the USAAF,claimed that,
in case of war with Japan, "Incendiary projectiles would burn the
[Japanese] cities to the groundin short order."38At a secret press conferenceon 15 November 1941, ArmyChief of StaffGeneral George C.
Marshall stated, "If war with the Japanese does come, we'll fightmercilessly. Flyingfortresseswill be dispatched immediatelyto set the paper
cities of Japan on fire.There won't be any hesitation about bombing
AirForce (Decimal File Number760.01,
Historyofthe Twentieth
36. NarratiNe
Sheets, Document 75,
1 July-2September1945), BinderVII, TargetInformation
HRA.
37. Crane,Bombs,Cities,and Civilians, 10, 120-36.
38. Sherry,Rise ofAmericanAirPower, 58.
MILITARY HISTORY
*
115
TIIOMAS R. SEARLE
civilians-it will be all-out."39Even withoutthe encouragementof such
influentialfiguresas Billy Mitchell and George Marshall,the Japanese
incendiarybombingof Chinese cities in the 1930s and the London blitz
of 1940 would have made it virtuallyimpossible forUSAAF planners to
have remainedunaware ofthe possibilityofusingfireto attackJapanese
cities.
USAAF planningin the Second WorldWarwas done in haste by inexperienced officersin newly created organizations.For the purpose of
choosing initial targets,the two most importantorganizationswere the
Air Staff,created in 1941, and the Committee of Operations Analysts
(COA), created in late 1942. The Air Staffserved all of the traditional
general stafffunctionsand was manned initiallyby a hardcore group of
airpowerenthusiasts,mostlyformerinstructorsat the Air Corps Tactical
School. It grew exponentiallyduring the war, but the planners were
lower-ranking,
less experienced,and altogetherless credible than their
counterpartsin the Armyand Navy.40The COA was a more ad hoc organization,created by General Arnoldto improvethe qualityand credibility of the intelligence on which USAAF planning was based. It was
composed primarilyof industrialand economic experts whose recommendations,Arnold hoped, would carry added weight with President
Roosevelt and Secretaryof War Henry L. Stimson. Much to the chagrin
of the Air Staff,the COA was independentof it and worked on special
projectsas requested by General Arnold.41Conflictbetweenthe Air Staff
and the COA was inevitable but was minimized by having prominent
membersof the Air Staffalso serve as membersof the COA.
Initially,both the Air Staffand the COA focused theirattentionon
the war in Europe. As a result,the Air Staffdid not startserious workon
selectingbombingtargetsin Japan untilFebruaryof 1943. On 24 February,the Chief of Operational Plans forthe Air Staffrequested "an overall targetstudy of Japan [and Japanese-controlledareas]" for an "air
offensiveagainstthe vital militaryand industrialcentersof the Japanese
Empire."42 The reply fromthe IntelligenceSection of the Air Staff,a
large binder entitled"Japanese TargetData-March 1943," consisted of
39. LarryI. Bland and Sharon RitenourStevens,eds., The Papers of George
Press,1981- ), 2: 678. The
CatlettMarshall (Baltimore:JohnsHopkinsUniversity
purpose of the secret press conferencewas presumablyto delivera threatto the
Japanesewithoutan officialpublicstatement.
40. Hansell,TheAir Plan thatDefeatedHitler(Atlanta,Ga.: Higgins-McArthur,
1972), 1-5. The AirCorpsTacticalSchool was theprewarheartoftheAirCorpsand
AirCorps doctrine.
and disseminating
was responsibleforformulating
41. Ibid., 148-49.
42. Japanese TargetData, March 1943, Air Force NumericalFile 142.621-1,
HRA.
116
*
THE JOURNAL OF
-_____________
"ItMade a LotofSensetoKillSkilledWorkers"
assessments of those Japanese industries that appeared to be worth
attacking.It assignednumbersto all targets,withsome identifiedas "key
targets"or "prioritytargets."However,in keepingwithUSAAF precision
bombing doctrine,no mention was made of area attacks on Japanese
cities. Two monthslater,in May of 1943, thatchanged. The Chiefof the
Plans Section requested, "as an addendum to 'Japanese Target DataMarch 1943' . . . a studyof the vulnerabilityof Japanese targetareas to
incendiaryattack."43
Withthisguidance,the IntelligenceSection preparedanotherbinder
entitled"Japan,IncendiaryAttackData, October 1943," which analyzed
twentykey cities and dividedeach into zones based on the flammability
of its structures.For the ten most importantcities, it provided overprintedmaps which indicated the locations of the various zones. The
most flammablezone of Tokyo was the targetof the 9 March 1945 raid.
The firstpage of the reportlisted fourreasons whyJapanese cities were
better targets than German cities for incendiary attack: the greater
ofJapanese residentialconstruction,the greaterbuilding
inflammability
congestion in Japanese cities, the proximityof factoriesand military
objectives to residentialconstructionin Japan,and the concentrationof
Japan'swar industryin a fewcities.44Thus, by October of 1943, almost
eighteenmonthsbeforethe 9 March raid, the Air Staffhad determined
that incendiary area attacks on Japanese cities would be dramatically
more effectivethan theyhad been against German cities.
Japanese civilian casualties figuredprominentlyin the minds of the
Air Stafffromthe beginningof theirplanningforincendiaryarea attacks
on Japanese cities. The Air Staffstudy"Japan,IncendiaryAttackData,
October 1943" listed the effectsthat the plannershoped to achieve with
incendiaryattacks. These were broken into "directeffects"(destruction
of production facilities,militaryestablishments,and storage facilities)
and "indirect effects" (reduced worker efficiency,casualties among
workers,damage to transportationfacilities,damage to public utilities,
diversion of resources to reconstruction, and lowered Japanese
morale).45Thus in 1943, USAAF plans forincendiarybombingofJapanese cities focused on the impact such attacks would have on Japanese
war production and included the clear intentto kill Japanese workers
and lower Japanese morale. As Arnold'swartimedeputy told a postwar
43. Japan,IncendiaryAttackData, October1943, coversheet,AirForceNumerical File 142.621-4,HRA.
44. Japan, IncendiaryAttackData, October 1943, Air Force NumericalFile
142.621-4,HRA.
45. Ibid.
MILITARY HISTORY
*
117
THOMAS R. SEARLE
interviewer:"It made a lot of sense to kill skilled workersby burning
whole areas."46
Shortlyafterthe IntelligenceSection of the Air Staffsubmittedits
initial report,the COA began a study of strategictargetsin Japan. It
chose to make its own independentstudybut had not yet assembled an
incendiarysubcommittee,so the COA report"Economic Objectives in
the Far East," dated 11 November1943, included the conclusions from
the recentlypublished"Japan,IncendiaryAttackData, October 1943."47
The COA reportlisted six criticalstrategictargetsystemsin the Japanese Empire withoutprioritizingthem or indicatinghow theymightbest
be attacked: merchant shipping,steel, urban industrialareas, aircraft
plants,antifriction
bearings,and electronics.To this list the JointIntelligence Committee of the JCS added petroleum (JIC 152/2). The JCS
approved these seven targetsystems on 6 April 1944, in JCS 742/6.48
Thus, in April of 1944-before the B-29 had flownits firstmission and
almost a fullyear beforethe Tokyo raid of 9 March 1945-the highest
U.S. military authorities approved Japanese urban areas as targets.
Henceforth,U.S. plans for bombing Japanese industrywould always
include both precision attacks on individualplants and area attacks on
cities.
The JCS also endorsed the Air Staffsinterestin Japanese civilian
casualties. A portionofJIC 152/2included in JCS 742/6explains the reason forattackingJapanese urban industrialareas in termsof the following intendedresults:
The absorption
ofmanhoursin repairand relief;thedislocationof
of publicservicesnecessaryto
laborby casualty;the interruption
offactories
production
and aboveall thedestruction
engagedin war
at pointsso
industrywould inevitablydisruptthe enemyeffort
a majordisaster[forJapan].49
numerousas to constitute
46. Eaker interview.General Eaker is best knownforhis wartimesenricein
Europe,but he was recalledto Washington
in March 1945 to serveas DeputyComwas to facilitatethe
manderoftheUSAAF.In thispositionhis primaryresponsibility
bombingofJapan.
47. Historyof the Committeeof OperationsAnalysts,16 November1942-10
October1944, vol. 1, 59-64, AirForceNumericalFile 118.01, HRA. One reasonfor
not establishinga subcommitteeto studyarea incendiarytargetsin Germanywas
thattheUSAAFhad access to thesubstantialBritishworkon theissue.
48. JIC 152/2and "Reportof Committeeof OperationsAnalystson Economic
Objectivesin the Far East" are bothincludedin JCS 742/6("OptimumUse, Timing,
and Deploymentof VeryLong Range Bombersin the War againstJapan") 6 April
1944, reel 1, frames0749-79,RecordsoftheJointChiefsofStaff,Part I: 1942-1945:
ThePacificTh-eater
PublicationsofAmerica,1981).
(Frederick,Md.: University
49. JCS 742/6,p. 48.
118
*
THE JOURNAL OF
"ItMade a LotofSensetoKillSkilledWorkers"
Thus the JCS in 1944, like the Air Staffin 1943, wanted to use civilian
casualties as a means of cuttingJapanese industrialproduction.Japanese morale, however,was not explicitlymentioned,indicatingthat for
the JCS, Japanese morale was not the main purpose of either the raids
or the civilian casualties.
Aftersubmittingits report on "Economic Objectives in the Far
East," the COA establishedan incendiarysubcommitteeand foundadditional supportforarea bombingJapanese cities. Beforethe war,Japanese industry relied very heavily on subcontractingto small "home
industries"scatteredthroughouthighlycongestedresidentialareas, and
theCOA assumed thatthishad not changed much duringthe war.50This
was not considered in earlier plans approved by the Air Staffand the
JCS, but wartimepress conferencesoftenmentionedthese tinymilitary
plants,widelydispersed in residentialareas. The COA feltthat incendiaryattacksshould not be conducted until a largeenough forcehad been
assembled to "give maximum assurance of totallydestroyingthe area
attacked,the danger being that a small effortwould merelycreate firebreaksagainsta laterheavy attack." The COA determinedthat,based on
historicalweatherdata, the best conditionsforstartinglarge fireswould
be in March and September.Its 9 May 1944 memorandumthereforerecommendedthat,ifan adequate forcewas available, a "generalattack on
Japanese urban industrialareas should be initiatedin Marchof 1945 and
concentratedduringthat month."751
Thus the COA, and throughit the Air Staff,recognizedthatincendiarybombingofJapanese cities was a complicatedundertaking.Ordinarilythe USAAF thoughtin termsof movingthroughits targetprioritiesin
a straightforward
way,destroyingthe highestprioritytargetsystemfirst,
followedby the second prioritytargetsystem,and so forth.But withfirebombing,small attacks might be counterproductive.Special weather
considerationsencouraged attackingcities in March regardlessof how
systems.The critmuch progresshad been made againsthigher-priority
how
many tons of
ical unanswered question facing the planners was
incendiarybombs would be needed per square mile to burndownJapan50. In fact,the postwarUnited States StrategicBombingSurveyfoundthat,
had been exaggerated.
exceptin Tokyo,the roleofhome industriesin the wareffort
the shiftto a war economy(aftertheUnitedStateswas deniedaccess to
Apparently
moved productioninto largerfirms.For a fuller
Japaneseindustrialinformation)
assessment,see UnitedStates StrategicBombingSurvey,The EffectsofAirAttacks
GPO, March1947), 29, 30. For targeting
on Japanese UrbanEconomy(Washington:
availableand thusthe perceived
purposes,the COA was limitedto the information
importanceofhome industrieswas critical.
51. Colonel Guido R. Perera (Chairman of the COA) to BrigadierGeneral
in "Historyof the COA 16 November1942-10
Hansell,9 May 1944, memorandum,
October1944," AirForceNumericalFile 118.01, HRA.
MILITARY HISTORY
*
119
THOMAS R. SEARLE
ese cities and, hence, how largea B-29 forcewould have to be assembled
beforemajor incendiaryraids could begin. Plentyof data were available
on the effectsof incendiarybombingon European cities, but the planners knew that Japanese cities were different.They hoped that small
experimentalraids would providethe needed informationwithoutdoing
too much to alert the Japanese. However,since "large"raids would have
dramaticallydifferentresults than "small" ones, the small test raids
would provideverylimitedinformationabout largeraids.52In retrospect,
it appears thatby changingtacticsand doublingthe bombload per plane,
LeMay transformedthe inadequate forceavailable to him in March into
a forcecapable of startingenormousfirestorms.
At this time, only the six cities where Japanese war industrywas
concentrated(Tokyo, Yokohama, Kawasaki, Nagoya, Kobe, and Osaka)
were thoughtworthyofattack,and U.S. plannersassumed thatthe same
reliefand recoveryservices would be used by all burned-outcities in
Japan. Attackingas many urban areas as possible in rapid succession
would help guarantee that Japanese recovery services were overwhelmed. Thus the COA saw urban area bombing as a brief and
extremelyviolent interludein the precision bombingcampaign against
Japan-an interlude that would begin when an adequate force was
assembled and end when the six key cities had been burned downregardlessof how the precisionbombingcampaign was going.The COA
hoped "to complete the planned destructionof all six cities within a
fromthe
periodofa fewweeks."53 Ideally "the entireMarch [1945] effort
Marianas" would be used againstthe six cities.54
Based on COA recommendations,the USAAF planned to begin
bombing Japan with small precision raids on critical factories,later
addingincendiaryarea attacksto the campaign.General Arnoldstuck to
this plan. For example, on 29 November 1944, Arnold's deputy commander of the TwentiethAir Force, BrigadierGeneral Lauris Norstad,
wrote Arnold a note suggestingan all-out attack against the Imperial
Palace in Tokyo with the small force available on 7 December 1944.55
Arnold'sreplywas, "Notat thistime,"and thatprecisionbombingshould
52. MinutesofMeetingofCOA, 13 September1944, p. 20, AirForceNumerical
File 118.151-16,HRA.
53. Minutesof COA Meeting,14 September1944, p. 25, Air Force Numerical
File 118.151-17,HRA.
forColonelLindsay,8 June1944,in HistoryoftheCOA,
54. COA Memorandum
AirForceNumericalFile 118.01, HRA.
55. This wouldhave been a classic "morale"raid since it wouldnotimpedethe
but mightdemoralizetheJapanese.Such a raidwouldalso have
Japanesewar effort
been well received by the U.S. public, and "morale" bombingcould be used to
improvefriendlymorale as well as demoralizethe enemy.Norstadto Arnold,29
November1944, letter,file373.2, OperationsReports,Aviation,RecordGroup 18,
D.C.
NationalArchives,'Washington,
120
*
THE JOURNAL OF
"ItMadea LotofSensetoKillSkilledWorkers"
continue: "Later destroythe whole city."56In his briefnote, Arnolddid
not specify exactly what would be different"later," but he clearly
intendedto wait untilhe had a large enough forceto "destroythe whole
city" and may also have intended to wait until more precision targets
had been destroyed.
The priorityassigned to urban industrialareas graduallyrose over
time but remainedfairlyacademic untillarge forcesarrived.In addition
to recommendingthat heavy incendiaryattacks be planned forMarch
1945, the 9 May 1944 memorandumto General Hansell also listedurban
targetsystem,behind the first-priindustrialareas as the third-priority
oritycoke plants (necessary for steel production) and the aircraftand
radio/radarindustriesthatshared second priority.In its finalreportof 10
October 1944, the COA raised the priorityof urban industrialareas to
second, behind only the aircraftindustry.57
Out on Guam, LeMay (who succeeded Hansell in January1945) and
his stafftried to execute the plans drawn up by the Air Staffand
approved by the JCS. The ordersforthe 9 March 1945 raid reflectedthe
longstandinginterestof the Air Staffand the JCS in using urban incendiary raids to cut Japanese industrial production by (among other
things) killingJapanese civilians. The Intelligence Section of LeMay's
staffproduced a targetinformationsheet for each mission, which the
bomber crews received and presumablyread. For the 9 March raid, the
targetinformationsheet was titled "Tokyo Urban IndustrialArea." On
page three, under the heading "Target Description," the information
sheet told the crews that "[w]ithinthis targetarea of approximately10
square miles, the average population density is 103,000 people per
square mile [hence over one millionpeople lived in the targetarea], an
averageprobablynot exceeded in any othermodernindustrialcityin the
world."Under "Importance,"the crews learned that:
of the targetarea wouldbe morenoticeablefromthe
Destruction
areaas a whole
pointofviewofitsrelationtotheTokyometropolitan
withinthe
thanfromthephysicalloss oftheindividualinstallation
notprofitable
targetarea itself.Howevernumeroussmallfactories,
attackwouldbe damagedordestroyed.
forprecisionbombing
targets
at scoresofwarplantsthroughout
Tokyoand environs
Employment
ofworkers
outof
wouldbe directly
affected
bycasualties,movement
and probablylowered
thearea,use ofmanpowerin reconstruction,
workermorale.58
his replyon Norstad'soriginalnote.
56. Ibid.; GeneralArnoldhand-wrote
57. "Historyof the COA, 16 November1942-10 October 1944," 1: 115, Air
ForceNumericalFile 118.01, HRA.
AirForce,BinderVII, TargetInformation
58. NarrativeHistoryoftheTwentieth
Sheets,Document75, DecimalFileNumber760.01, 1 July-2September1945,vol. 8,
HRA.
MILITARY HISTORY
*
121
THOMAS R. SEARLE
Thus, casualties were again explicitlymentioned,and regardedas desirable because theywould directlyaffectemploymentat war plants. This
time, however,the planners sought not just casualties among workers
but also casualties among those (like the membersof the workers'families) whose injuries or deaths mightkeep workersfromgoing to work.
Their casualties were importantonly inasmuch as they contributedto
the general dislocationin Tokyothat the plannershoped to achieve as a
means of disruptingwar production.The planners also hoped to lower
Japanese "worker morale" but, unlike casualties, which the planners
were sure the raid would produce,lower moralewas merelya "probable"
outcome ofthe raid. Loweringmorale was a means ofloweringindustrial
production-not a means ofincitingeitherrebellionor populardemands
forsurrender.
The planningforthe 9 March raid was unusually thoroughand the
targetinformationsheet was unusually detailed. The targetinformation sheets were typicallyonly a page or two long, and very briefon
what the attack might achieve. They often neglected casualties,
morale, and all the other "indirect effects" the raids hoped to
achieve.59 But the USAAF's internal wartime propaganda vehicle,
Impact magazine, provided detailed articles on the fire raids that
repeated the planners' view that casualties were one of the ways area
raids cut Japanese production.60Everyone was well aware that burning down vast tracts of Japanese cities would produce substantial
Japanese civilian casualties. For example, a front-pagearticle on the
firebombingappeared in the New York Times under the sub-headline:
"1,000,000 Japanese Are Believed to Have Perished in Fires."'61 Few
Americans complained and many may have welcomed these Japanese
casualties.62
59. This is not to say thattheTokyosheetwas unique.Othertargetsheetsmention "casualties," "morale,"and "absenteeism,"but the consistentfocus was on
Japaneseindustry.
The targetinformation
sheetscan be foundin NarrativeHistoryof
the Twentieth
Air Force,BinderVII, TargetInformation
Sheets,Document75, Decimal File Number760.01, 1 July-2September1945, vol. 8, HRA.
60. See forexample:"AreaBombingWrecksJap'Home'Industrv,"
Impact,April
1945; and "FireBlitz:ProgressReporton theIncendiaryBombingofJapan,"Impact,
to the first(April1943) issue ofImpact, he
August1945. In Arnold'sintroduction
explains that the purpose of the magazine is to "bringhome to our Air Forces
the worldand to our comradesin otherareas of the servicewhat the
throughout
force,or impact,ofair poweris and can be." The entirerunofImpact was reprinted:
AirForceHistoricalFoundation,Impact (New York:J. Parton,1980).
61. The fullheadlinereads:"51 Square MilesBurnedOut In Six B-29Attackson
Tokyo,LeMayBacks FiguresWithPhotosofHavoc-1,000,000 JapaneseAreBelieved
to Have Perished,"New YorkTimes,30 May 1945.
62. JohnWV.
Dower'sWar WithoutMercy:Race and Power in thePacific War
(New York:Pantheon,1986) providesan excellentassessmentof wartimeracismin
122
*
THE JOURNAL OF
"ItMade a LotofSensetoKillSkilledWorkers"
Afterthe war, Arnold retained his enthusiasm for using Japanese
civilian casualties to impede wartimeindustrialproduction.For example, writingas head of the USAAF in his final"WarReport,"he describes
the firebombing
ofJapanese cities in detail and concludes that:
In additionto thedestruction
ofindustrial
installations,
thecasualties caused had significant
effects
in dislocationof industrial
manpowerand on enemymorale.The Japanesehave statedthatair
attackskilled260,000,injured412,000,left9,200,000homeless,
and demolished
orburneddown2,210,000houses.63
Michael S. Sherrymisrepresentsthis situationin his award-winning
book, The Rise ofAmerican Air Power. Sherryclaims that "the vague
circumlocutions employed and the incremental way by which new
assumptions crept into planning obscured the shift[in targetpriority
toward urban areas]."64 But the 9 May 1944 memorandum Hansell
received fromthe COA clearly made urban areas high-priority
targets
and recommendedthat theybe accorded the highestpriorityin March.
The COA was not beingvague or circumlocutoryin explicitlyraisingthe
priorityof attacks on urban areas. For theirpart,the planners working
forLeMay in the Marianas were quite explicit in their intentionto kill
Japanese civilians,and the Air Staffand the JCS had not been vague in
favoringincendiaryarea attacks and explicitlylistingcivilian casualties
as one of the goals of incendiaryattacks. If "new assumptionscrept into
planning,"theydid so back in May of 1943 when the Chief of the Plans
Section of the Air Staffrequested an incendiarystudy.
Changes of Command: Or, Why Hansell Did Not Matter
BrigadierGeneral Haywood S. Hansell was a staunch advocate of
precision bombingand perhaps fanaticalin his commitmentto the idea
that destroyinga relativelysmall numberof carefullyselected factories
was the most efficientway to win the war.65Since the major area raids
bothJapanand the UnitedStates.Dowerdoes not claim,and thedocumentsdo not
indicate,that U.S. firebombing
was the resultof U.S. racism,but hostilityto the
Japanesecertainlycolored the generalpublic'sview of the bombing,and decision
makerswereaware thattheywouldreceivelittlecriticismforJapanesecasualties.
63. The fulltitleofthereportwas "ThirdReportoftheCommandingGeneralof
theArmyAirForcesto the SecretaryofWar,"whichcan be foundin GeorgeC. Marshallet al., The WarReportsofGeneraloftheArmyGeorgeC. Marshall,Generalof
theArmyH. H. Arnold,and FleetAdmiralErnestJ. King(Philadelphia:J.B. LippincottCo., 1947), 440. These war reportswereunclassifieddocumentspublishedduringand immediately
afterthewar.
64. Sherry,Rise ofAmericanAirPower, 258.
65. Hansell,Germanyand Japan, 211.
MILITARY HISTORY
*
123
TI-IOMASR. SEARLE
againstJapan began shortlyafterhe was relievedof command,it is natural to wonderifhis oppositionto such raids was the reason forhis relief.
Afterthe war,Hansell claimed thatcriticalchanges were made to the
plan afterhe leftthe staffand became an operational commander.66He
also implied that he would not have conducted extensive area bombing
if he had remained in command and that his removal representeda
change in policy away fromprecision bombingand towardarea bombing. Some historianshave made these same claims explicit. Sherry,for
example, makes much of the factthatHansell leftWashingtonbeforethe
finalCOA reportcame out and thus "had missed out on a crucial stage
in planning." Sherry also implies that Hansell was fired because he
opposed area bombingand that his replacementby LeMay represented
a change of policy in the bombingof Japan.67
The finalCOA reportdid in fact come out afterHansell leftWashington,and the reportpromotedJapanese urban areas to the second
most importanttargetsystem,ahead of everythingexcept the Japanese
aircraftindustry.This would seem to supportHansell's claim,but Hansell
received the earlier COA reports,and he had seen urban areas steadily
rise on the targetprioritylist. Though he did not reviewor endorse the
finalCOA report,he did receive the 9 May reportthatmade urban areas
the thirdprioritybehind the coke ovens (firstpriority)and the aircraft
and radio/radarindustries(sharing second priority).Since virtuallyall
the coke ovens were beyond the range of Hansell's planes in the Marianas, and he never attacked them, the previous reporthad effectively
made urban areas second priorityfor him. The only "critical stage in
planning"Hansell missed was the ratherminor decision that Japanese
urban areas were better targets than the three key plants in the
radio/radarindustry.68
IfHansell had stayedin Washingtonlong enough to quarrel withthe
finalCOA report,he certainlywould not have removedJapanese urban
areas fromthe targetlist because he did not object to includingthem.69
The most Hansell could have done would have been to keep urban areas
from moving up the prioritylist, that is, keeping them behind the
radio/radarindustry.But the 9 May 1944 reportHansell receivedexplicitly recommended a pause in the precision campaign to conduct area
bombingof the six key cities (ideally in March) ratherthan a lock-step
66. Hansell,StrategicAir WarAgainstJapan, 50, 51.
67. Sherry,Rise ofAmericanAirPower, 258.
68. Historyof the COA, MemorandumforBrigadierGeneral Hansell,9 May
1944, AirForce NumericalFile 118.01, HRA.
69. Hansell plannedto destroyurban areas (and kill largenumbersof enemy
civilians)in bothGermanyand Japanbut feltthattheyshouldbe attackedonlyas a
"last resort,"afterdestructionof all precisiontargetshad failed to convince the
enemyto surrender.(Hansell,Germanyand Japan, 47, 216, 217.)
124
*
THE JOURNAL OF
"ItMade a LotofSensetoKillSkilledWorkers"
progressionthroughthe targetpriorities.Whethertherewas one industryor two ahead of urban areas on the targetlist was not the issue.
Even in an organizationas closely watched by Washingtonas the
TwentiethAir Force, operationalcommandershad some latitudein how
theyconducted theirtrainingand operations.One of the lessons of the
bombing of Germany seemed to be that small forces of unescorted
bombers could not surviveover the enemy's homeland in daylight.The
factorieswould eventuallysupply the USAAF with large numbers of B29s and the capture of Iwo Jimawould eventuallygive the B-29s fighter
escort over Japan,but initiallythe B-29s would be veryvulnerable.The
firstcommanderof operational B-29 forces,General Wolfe,took this to
heart and emphasized nightradar bombingratherthan daylightvisual
Not too surprisingly,
Wolfe'sforcewas ineffective
bombing.70
against the
precision targetsthe plannersgave him and when Arnoldreplaced him
withLeMay,the firstthingLeMay did was retrainthem to improvetheir
daylight accuracy.71 The men Hansell commanded (mainly the 73d
Bomb Wing) were originallyslated to join Wolfein India, and theirtraininghad also emphasized nightradarbombing.Like LeMay,Hansell's first
move as commanderwas to trainhis crews fordaylightprecision bombing.72Thus, nightradarbombingwas a partof B-29 operationsand training from the outset. Wolfe favored these operations because they
increased the safetyof his crews, and both LeMay and Hansell independently moved away fromthese tactics toward the improvedaccuracy
provided by daylightoperations. However,the crews and planes were
ready foreithertypeof operation.
The Twentieth Air Force Headquarters in Washington pushed
Hansell to conduct experimentalurban incendiaryattacks in December
of 1944, and he protestedthis pressure. But Hansell was under a lot of
strain about many thingsand protestedoften.73Afterthe war, Hansell
found himself in the uncomfortableposition of having vehemently
opposed the two mosteffectiveaspects of the B-29 campaign:aerial min70. Ibid., 165, 167. Earlyin the B-29 crew trainingprogram,veryfewaircraft
were available,so it was easier to traincrewsin radaroperations-whichcould be
done in individualaircraft-rather
thanin formation
flying,
whichrequireda group
of aircraft.This is not to say thatthe crewsweregood at radarbombing.Whenhe
took over,LeMay was appalled at how inadequate the radar operatorswere,as he
recallsin LeMayand Kantor,
Mission withLeMay,345, 346.
71. Hansell,Germanyand Japan, 209.
72. Ibid., 170.
73. In his memoirs,HansellacknowledgesArnold'sinsistencethathe improve
bothmaintenanceand operationsand goes on to lamentArnold'sobsessionwithstatisticsand theconstantstreamofmessagescomingintohis headquartersdemanding
information
he did not have. For Hansell'sview,see his StrategicAir WarAgainst
Japan, 44, 48; and Germanyand Japan, 162.
MILITARY HISTORY
*
125
TIOMAS R. SEARLE
ing of Japanese waters and urban incendiary bombing.74He opposed
bothbecause theywere distractinghim fromhis precisionattackson the
Japanese aircraftindustry.His belief that these were requests, or suggestions,ratherthan ordersfromGeneral Arnold,increased the vigorof
In the case ofaerial mining,he was essentiallycorrectthat
his protests.75
this was the Navy's idea and that Arnold was participatingreluctantly
and was generallyin sympathywithHansell's views. In the case of area
bombing,Hansell thoughthe was only arguingwith General Norstad
(Hansell's successor as chiefof staffof TwentiethAir Force) and thathe
could win the argumentby appealing to Arnold.Arnold,however,agreed
with Norstad to a much greater degree than Hansell realized; when
Hansell lost his appeal to Arnold,he conducted an experimentalarea
incendiaryraid on 3 January1945 as ordered.76Afterthe war,Hansell
was surprised to learn that General Arnold had long supported area
bombingof Japan.77For all his complaining,Hansell never even hinted
at resigningor disobeyingorders. If area bombingwas what it took to
stay in command, Hansell would have continued to do as much of it as
Arnoldtold him to.78
The counter-argument
to the claim thatHansell would not have conducted area bombingofJapanese cities is that much of his bombingwas
just that. There was the 3 January1945 raid noted above but, fromhis
firstraid on Japan on 24 November 1944, Hansell aimed a significant
B-29s
portionofhis bombingat urban areas. On thatfirstraid,thirty-five
bombed the aircraftfactory(the primarytarget),but fifty
B-29s bombed
the urban area of Tokyo (the secondary target).79The next raid on 27
Novembersaw no B-29s bomb the primary(aircraftfactory)targetand
74. Of course the best-knownB-29 raids were the two atomic bombings,but
theywerea part of the atomicprogramthatjust happenedto use a fewB-29s and
werenot reallya partof the TwentiethAirForce'scampaign.For Hansell'sviewson
aerialmining,see Hansell,Germanyand Japan, 198-201.
75. Ibid.,199, 218.
76. Hansellneglectsto mentionthismissionin his memoirsbutit can be found
in, TwentiethAir Force: A StatisticalSummaryof Its OperationsAgainstJapan,
NumericalFile 760.308 (June44-August45), HRA; and Kit C. Carterand Robert
Mueller,The ArmyAir Forces in WorldWar II: Combat Chronology,1941-1945
OfficeofAirForce History,1973), 538.
(WVashington:
77. Hansell,Germanyand Japan, 216-19.
78. One of the strikingthingsabout Hansell'smemoirsis that,wIhilehe often
criticizesArnold,he never mentionsan instance when he openly disagreedwith
It is veryhardto escape theimpresArnoldor changedArnold'smindaboutanything.
sion thatArnoldlikedHansellbecause Hanselldid exactlywhatArnoldtoldhim to
do; and thatHansellwouldhave continuedto do Amold'sbiddingin usingB-29s for
area incendiarybombing.
79. MissionSummary,
MissionNumber7, SummaryofXXI BomComMissions,
760.01, vol. 6, 1 July-2September1945, HRA.
126
*
THE JOURNAL OF
-________________________
--"It
Madea LotofSensetoKillSkilledWorkers"
fifty-nine
bomb the urban secondary target.80Hansell's third raid on
Japan (29 November) was a straightforward
nightarea incendiaryraid
on Tokyo.8' Thus, Hansell conducted a night area incendiaryraid on
Tokyobeforehe was pressuredto do so, and even when he planned a precision raid on a factory,it oftenbecame an area raid on a city.
The companion error to exaggeratingthe importance of Hansell's
dismissal is to misunderstandLeMay's role.82UnlikeHansell, LeMay was
not involvedin planningthe bombingofJapan; at the time he was commandinglarge bomber forcesin Europe. There he conducted precision
bombingof factories,as well as radar bombingof urban areas. Contrary
to the claims of Ronald Schafferand other historians,LeMay did not
arrivein the Marianas focused on incendiaryarea bombing.83The firebombinggets considerable attentionin LeMay's memoirs and postwar
interviewsbecause of its spectacular success, and after9 March 1945,
LeMay ordereda lot of it.84But in India, China, and the Marianas before
March 1945, he focused on precision bombing.As he says in his memoirs and repeatedin interviews,his firstmajor changes upon takingcommand in India, China, and the Marianas were to improve lead-crew
training,formationflying,and other aspects of precision bombing.85
Before9 March his operations out of China and the Marianas were the
same mix of many precision raids and a few area raids employed by
Hansell. Only twoofHansell's firstnine missionsagainstJapanwere area
raids and only twoof LeMay's firsteightmissionsagainstJapan fromthe
Marianas were area raids.86
Even afterthe 9 March raid, LeMay had no intentionof abandoning
precision bombingand in fact did as much as the weather allowed. In
Julythe weather was verybad and only 16 percent of his sorties were
precision raids, but betterweatherenabled him to devote 27 percent of
80. MissionSummary,MissionNumber8, SummaryofXXI BomComMissions,
ibid.
81. MissionSummary,MissionNumber9, SummaryofXXI BomComMissions,
ibid.
82. For example,Crane, Bombs, Cities, and Civilians, suggeststhat it was
LeMay'swillingnessto abandon precisionbombingthatdistinguished
the USAAF's
bombingcampaignagainstJapanfromthatagainstGermany.
83. Schaffer,
WingsofJudgment,125.
84. AfterMarch1945, in responseto the spectacularsuccess ofarea incendiary
bombing,theUSAAFexpandedthenumberofcitiesconsideredworthy
ofattackfrom
theoriginalsix to overseventy.Fordetails,see Cravenand Cate,ArmyAirForces,5:
653-58.
85. Kohn and Harahan,StrategicAir Warfare,55-58, 61, 62; and LeMay and
Kantor,
Mission withLeMay,328-45.
86. Twentieth
AirForce,A StatisticalSummaryofIts OperationsAgainstJapan,
NumericalFile 760.308 (June1944-August1945), HRA.
MILITARY HISTORY
*
127
TI-IOMASR. SEARLE
his June sorties to precision bombing,and 37 percent of August's.87He
even attemptedto beat the weather by using low-altitudenightattacks
against precision targets.88Thus neitherLeMay nor the USAAF abandoned precision bombinginfavor of area bombing.Instead, theysupplemented an unspectacular precision bombing campaign with a
stunninglysuccessful urban incendiary campaign. (In Europe, on the
other hand, the USAAF's area bombingcampaign was relativelyunimpressive compared to its precision campaign or the Britisharea campaign.) LeMay recognized that oftenthe weatherwould not permithim
to attack his highestprioritytargets,so he founda way to take advantage ofweaknesses in Japanese defensesand to devastatesecondary targets when the weather was bad.89Urban incendiarybombing was not
LeMay's idea, or even his primarygoal; he just made it work spectacularlywell.
The differencebetween Hansell and LeMay is illustratedby the differencebetween the raid conducted by Hansell on 27 November 1944
and the one conducted by LeMay on 25 February 1945. Both men
wanted to attack aircraftfactories.Both men wound up attackingurban
areas. But Hansell sent out his planes prepared to attack factoriesand
made a haphazard attackon an urban area; LeMay postponedhis factory
attack and sent out his planes preparedto conduct the urban area attack
theyexecuted. As LeMay's staffwritesin the reporton the 25 February
mission, the primaryvisual targetwas the Musashino aircraftengine
plant near Tokyo, but "weather forecastsindicated that all of Honshu
[the island on which Tokyo is located] would be overcast,necessitating
the selection of a radar target.As a resultthe urban area of Tokyo was
chosen."90
The "hiring"of LeMay and the "firing"of Hansell needs to be understood not in terms of area versus precision bombingbut ratherwithin
the contextofArnold'sconcept ofhow the B-29 campaignwould develop
and who his subordinate commanders would be. As noted above, two
critical elements of Arnold'sargumentforhis personal control of B-29
operationswere the factthatthe planes would be directedagainstJapan
(rather than targetswithinone of the theaters) and that theywould be
based in more than one theater.The firstcriterionwas met by rejecting
out of hand all but the most desperate requests forsupportfromtheater
87. Ibid.; and, UnitedStatesStrategicBombingSurvey,The StrategicAir Operation of VeryHeavy Bombardmentin the WarAgainstJapan (Washington:GPO,
1946), 16-17.
88. Cravenand Cate,ArmyAirForces, 5: 645-53. These methodsfailedto produce dramaticimprovements
in bombingaccuracy.
89. As LeMaysaid,he was "Tryingto getus independentofweather"[emphasis
LeMay's],in LeMayand Kantor,MissionwithLeMay,351.
90. ReportofOperations,25 February1945, Box B26, LeMayPapers.
128
*
THE JOURNAL OF
"ItMade a LotofSensetoKillSkilledWorkers"
-_________________________
commanders. To achieve the second, operations had to be conducted
fromseveral theaters.9'Operations out of China were immediatelypossible and would become more attractiveif the Allies could reestablish
overland communicationsbetween India and China and push back the
Japanese in China, makingbases available closer to Japan.The Marianas
would be the second B-29 base area to become available, and the Philippines would be the third.The Marianas were the most promisingsince
operations therecould begin relativelysoon (unlike in the Philippines),
and its bases would not depend on the offensivecapabilities of Chiang
Kai-shek'sforces.WithChina in the China-Burma-Indiatheater,the Marianas in the CentralPacifictheater,and the Philippinesin the Southwest
Pacifictheater,this schedule met the need to base out of all threemajor
theaters. Arnold created three Bomber Commands to conduct these
operations:the XX forChina, the XXI forthe Marianas, and the XXII for
the Philippines.92
Arnold faced a problem findinggood commanders for B-29 operations because his best bomber commanderswere in Europe. As soon as
the Germans surrendered,Arnold sent his top officersfromEurope to
take over the campaign against Japan.93Until then, he had to make do
withmorejunior officers.His plan was to give the XX BomberCommand
in China to Wolfe,a brigadiergeneral; the XXI in the Marianas to LeMay,
a major general; and the XXII in the Philippinesto Hansell, a brigadier
general.94While Arnolddid not leave a record of his reasons formaking
these selections,it is worthnotingthat these threeofficersrepresented
key constituencieswithinthe USAAF: Wolfecame fromthe production
and logisticsside, LeMay spent the war commandingoperationalbomber
units, and Hansell was best known as a planner and staffofficer.The
sequencing of theirassignmentsmay have been due to the factthat the
greatest challenges to early operations would be mechanical, which
Wolfewas extremelywell prepared to handle. The operationsout of the
Marianas would be the USAAF's firstreal chance to put largenumbersof
bombs on key targets,and thatcommand would be much largerthan the
others. LeMay had shown a remarkableabilityto put bombs on targets,
91. Hansell,StrategicAir WarAgainstJapan, 26.
92. Hansell,Germanyand Japan, 164.
93. Gen. Carl A. Spaatz took over directionof the campaignfromAmold; Lt.
Gen. JamesH. Doolittlebroughtin his EighthAirForce fromEngland;and Lt. Gen.
NathanF. TwiningleftItaly to replace LeMay as commanderof the TwentiethAir
Air Force in
Force. (In July1945, Arnoldhad closed the headquartersof Twentieth
Washington
and redesignedthe XXI BomberCommandas the TwentiethAir Force
underLeMay'scommand.)No one explainedthe changein commandto LeMayuntil
Twiningarrivedand LeMay asked him "Whatare you doinghere?" UnlikeHansell,
LeMaychose to stayon afterlosinghis command.For relateddiscussion,see Hansell,
StrategicAir WarAgainstJapan, 69-7 1.
94. Ibid.,45.
MILITARY HISTORY
*
129
TIOMAS R. SEARLE
and he had much more experience commandinglarge bomber forces
than had Wolfe and Hansell combined. By the time the Philippines
became available forB-29 operations,the TwentiethAir Force would be
runningsmoothlyand Hansell's job as chief of staffwould be winding
down.
Regardless of Arnold's reasons for choosing Wolfe, LeMay, and
Hansell to be his firstthreeB-29 commanders,eventsled him to modify
his originalplan. When the XX Bomber Command's operations disappointedhim,Arnoldreplaced WolfewithLeMay. This moved Hansell up
from commanding the XXII to the XXI, in the Marianas. WVhenthe
USAAF decided not to use the Philippines as a B-29 base, the XXII
Bomber Command was disbanded. Thus the only reason Hansell ever
commanded any B-29s was that Wolfe was fired.When the Japanese
ground offensivedrove the XX Bomber Command out of China and
Arnold decided to consolidate all B-29 operations in the Marianas, he
had another command problem.95AllowingHansell and LeMay to run
competingcommands out of the Marianas did not make sense, so Arnold
had to choose between them.This was a veryeasy choice because there
were compellingreasons forkeepingLeMay and replacingHansell.
The first,and most obvious, factwas that LeMay outrankedHansell
in 1945; the simplestsolution,then,was to make LeMay the commander with Hansell as his vice commander.This was in fact what Arnold
did, but Hansell quit (in somethingof a huff)ratherthan serve under
LeMay.96Aside fromhis rank,LeMay was also the bettercommander.Of
all Hansell's many defenders,none claims that Hansell would have done
a betterjob than LeMay did.97Arnold'soriginalplan forwho would get
which command suggests that, before operations began, he preferred
LeMay to Hansell and, had the plan called foronly one B-29 base area,
he would have given that command to LeMay and not to Wolfe or
Hansell.
WhateverArnold'spreferencesbeforeoperationsstarted,by January
1945 he had the benefitof several monthsobservingLeMay and Hansell
95. Thoughthe Japanesedid not actuallycapturethe airfields,theiroffensive
put such a strain on Allied logisticsin China that the U.S. commanderthere
requestedthatthe B-29s be removed.For relateddiscussion,see Cravenand Cate,
ArmyAirForces, 5: 150-52.
afterquitting,see
96. For one ofthewhineylettersHansellsent to Washington
Hansell,StrategicAir WarAgainstJapan, 140-43.
97. The closestanyonecomes to claimingthatHansellwas the equal ofLeMay
is Hansell'sown descriptionofhow,plannedwiththe fullbenefitof hindsightand a
unavailableduringthewar,a precisionbombingcampaign(prewealthofinformation
sumablyconductedby him) mighthave done just as well as the campaignLeMay
actuallyran,butwouldhave takenlonger.For thisdiscussion,see Hansell,Strategic
Air WarAgainstJapan, 71-93.
130
*
THE JOURNAL OF
"ItMade a LotofSensetoKillSkilledWorkers"
--__________________________
runningindependentB-29 operations(LeMay in India since August1944
and Hansell in the Marianas since October 1944). Arnoldmade his view
of their respective effortsclear when he chose to put Hansell under
LeMay's command and leftno othercommenton the issue. Others have
not been so reticent.Arnold's chief of staff,General Norstad,feltthat
Hansell sufferedfroman "utterabsolute complete and irreversiblelack
of competence."98One Air Force analyst pointed out the difference
between the reportscoming into Washingtonfromthe two B-29 commanders: "LeMay was writinghalf-pagereportstellingArnold what he
did yesterday,and Hansell was writinga three-pagereportexplaining
why the mission aborted."99This disparitywould have struckeven the
most casual observerbut, given that Arnoldwas demandingaction and
hated to receive reportsthat were more than one page long, Hansell's
catalogue of excuses only served to weaken furtherhis untenable position. The factthat he wrotesuch reportsat all indicates that he had no
idea how his actions would appear to others and completelymisunderstood his position.
Aside fromthe enormous advantages LeMay had over Hansell, there
were solid reasons for simply firingHansell. The ineffectivenessof
Hansell's operations has already been noted. In addition, his subordinates and some of his superiors intensely disliked him. As Hansell
acknowledged,his main combat force,the 73d Bomb Wing,"was openly
Hansell also
hostile to me," fromits commanderto its lowest private.100
had a runningfightwithMajor General WillisH. Hale, the commanderof
the other USAAF aircraftin and around the Marianas.101To make mattersworse forHansell,he managed to alienate thoroughlya groupofcongressmen who visited him on the Marianas.102 Arnold got official
complaints about Hansell fromHale and the congressmen.Arnold also
knewthatHansell's subordinateshated him because the key subordinate
commander (BrigadierGeneral Emmett "Rosey" O'Donnell, Jr.,of the
73d Bomb Wing) was a good friendofArnold'sand was doingeverything
98. Interviewof Brig.Gen. Lauris Norstadby MurrayGreen, 15 July1969,
microfilm
168.7326, roll43825, p. 14, HRA.
99. Interviewof Maj. Gen. JohnB. Montgomery
by MurrayGreen, 8 August
1974, microfilm
168.7326, roll43825, p. 9, HRA.
100. Brig.Gen. H. S. Hansellto Lt. Gen. BarneyM. Giles,27 March1945, letter,
reprintedin Hansell,StrategicAir WarAgainstJapan, 140-43. Whileit is not clear
whyeveryonein the 73d BombWinghated him,Hanselldoes describea clash with
BrigadierGeneral EmmettO'Donnell, Jr.,over theirfirstmission against Japan.
O'Donnell feltthat the daylightraid Hansell insistedon was too dangerousand
insteadfavoreda night(area) raid.For a discussionofthis,see Hansell,StrategicAir
WarAgainstJapan, 37, 38.
101. Hansell,Germanyand Japan, 208.
102. Ibid., 185.
MILITARY HISTORY
*
131
THOMAS R. SEARLE
he could to get Hansell fired.103
Withthe open and outspokenhostilityof
so many people and without operational successes to fall back on,
Hansell was not likelyto keep his job long.
Even his personal friendsfeltHansell was not the man forthe job.
Several key membersof the Air Staff,includingLieutenantGeneral Barney Giles and Major General Lawrence Kuter, actually approached
Arnoldto request thatHansell not be givencommand ofthe XXI Bomber
Command. Giles and KuterknewHansell as well as anyone and were two
of Hansell's closest friends,but theyhad no faithin him as commander
in the Marianas. A postwarinterviewersuggestedto General Giles that
Hansell had not been givenenough time,but Giles passed up the chance
to defendhis old friendand said, "No, no, he was the wrongman to send
there in the firstplace."'104How long Arnold mighthave put up with
Hansell is not clear,but when the opportunitycame to move him out of
command and back into a staffposition (under LeMay), it must have
been veryattractive.
Hansell's biographerrefersto Hansell's firingas a personal "tragedy,"
but the tragedywas Hansell's narcissismand completelack ofself-awareness.'05A littlethoughton Hansell's partwould have revealed to him that
he would have a commandonlyiftherewere a B-29 command thatWolfe
and LeMay could not hold. It was also clear he would lose thatcommand
ifGeneral Arnoldconsolidated B-29 operations.And any of the threeof
themwho was stillaround would be pushed aside the momentthe European commanderswere available forPacific duty.As it turnedout, the
command Hansell was originallysupposed to get (XXII Bomber Command) was disbanded and never saw action. LuckilyforHansell, WVolfe
failedbeforeLeMay could take over XXI Bomber Command in the Marianas, so Hansell got XXI (the best B-29 job) untilArnolddecided to stop
wastinghis time in China. Then LeMay got the command Arnold had
always intended to give him and kept it until the USAAF's firststring
arrivedfromEurope. For the rest of his life, Hansell puzzled over his
relieffromduty with the same lack of self-awarenessthat he showed
when he responded to his reliefby sendingArnolda ten-pageletterthat
said in part,"I feel,on reflection,that I have erred in not passing on to
you my problemsin more detail."1106(How could he have imaginedthat
103. Interviewof Maj. Gen. JohnB. Montgomery
by MurrayGreen,8 August
1974, p. 7, microfilm
168.7326,roll43825, HRA.
104. USAF Oral HistoryProgram,interiew with Lt. Gen. BarneyM. Giles,
20-21 November1974,pp. 91-93, K239.0512-814,HRA.
105. Griffith,
The Quest. Griffith
claims thatHansell was firedbecause of his
reluctanceto conductarea bombing.
106. Hansell to Arnold,14 January1945, RecordGroup 18, File 201, National
Archives(quoted in Griffith,
The Quest, 196).
132
*
THE JOURNAL OF
"ItMadea LotofSensetoKillSkilledWorkers"
Arnoldwould read a ten-pageletterfroma man he just fired?) If ever a
man complained enough,it was Haywood Hansell.
So wheredoes all thisleave us? The 9 March 1945 raid on Tokyowas
a radical break frompreviousoperationsbecause LeMay had the imagination and courage to trylow-altitudenightoperations,not because it
was an incendiaryraid on a major city.In bad weather,the USAAF performedextensive area incendiarybombingof both German and Japanese cities. From 1943 on, USAAF plans included incendiary area
bombingof major Japanese cities, and heavy attacks were supposed to
startwell afterthe precision bombingcampaign began. The main target
of the raids was Japanese industrialproduction,and one of the means
used to cut that productionwas civilian casualties. The intentionto kill
large numbers of Japanese civilians was explicitlyincluded in planning
-documentsread and approved at every level fromthe individual aircrewman to the JointChiefsof Staff,and in his finalreporton how he
conducted the air war,the CommandingGeneral of the USAAF included
heavy Japanese civiliancasualties as a measure of his success.
MILITARY HISTORY
*
133