Chapter 2

The Casablanca Offensive: The Allies over Germany, 1943–44

At lunchtime on January 18, 1943, Air Vice Marshal John Slessor, RAF assistant chief of staff, sat on top of the roof of the Anfa Hotel in Casablanca watching “the long Atlantic rollers breaking on the beaches” while he sketched out a compromise agreement between the American and British chiefs of staff over the future of Allied strategy. Chief of Staff Charles Portal then read it through and changed a few words. In the list of strategic commitments jotted in his notebook, Slessor had included “The heaviest possible bomber offensive from the UK against GERMANY direct.”¹ His hastily concocted notes were typed up and agreed to when the Combined Chiefs of Staff reassembled for the afternoon session, and they became the basis for the document on Allied strategy endorsed by Roosevelt and Churchill three days later. Slessor elaborated the idea of a heavy bomber offensive into a full draft directive, and this was presented to the Combined Chiefs on January 21 with only minor changes in the wording. It was approved, and the Casablanca Directive for a joint bomber offensive against Germany was released as policy document CCS 166 two days later.²

The Casablanca Conference (January 14–24) came at a critical point for the Allies. Stalin declined to come, being too occupied with the battle for Stalingrad, so the discussions focused on the future of Western Allied strategy. At stake was the balance between expanding the Mediterranean theater of war, which the United States had joined with the landings of Operation Torch in November 1942, and the plan to open a second front in France in 1943 or 1944. For the bomber forces there was more at stake. The conference opened at the end of a period of growing criticism of Bomber Command and the Eighth Air Force; it presented both forces with the opportunity to argue their case for sticking with an independent bombing strategy. This entailed a public relations exercise to sell bombing to a potentially skeptical audience. General Arnold instructed his staff to prepare detailed statistics, maps, reports, and colored charts for him to take to Casablanca, a list of props that ran to over three pages.³ Harris took pains to ensure that a regular flow of publicity material, including good aerial photographs of damaged cities, reached the American press. The Air Ministry organized an exhibition in Washington in early January 1943, which was visited by Vice President Henry Wallace and later taken to the White House to show to Roosevelt. Wallace, it was reported, was “completely sold on the necessity of bombing Germany” as a result of what he had seen, and keen to pass on his impressions to the president. The RAF delegation in America thought of producing a film to assist Arnold’s efforts to present the bombing offensive to the American public.⁴ At Casablanca, Arnold fielded a full team with both Eaker and Spaatz in attendance to argue the air force case; Portal had chosen to take Slessor, a sociable air force diplomat with planning experience, rather than Harris, whose bluntness would have been out of place in the delicate discussions to follow.

It is difficult to assess whether these propaganda efforts really affected the final decision to approve a combined offensive. The outcome for the two bomber forces was in the end mixed. The Casablanca Directive was a loosely worded document, a set of hopeful intentions rather than a clear plan “that could have been made by any schoolboy,” as one senior RAF officer later put it.⁵ Months went by before a real planning document was produced. It was also designed to fit in with the priorities of the other services and the political leadership. Bombing was accepted at Casablanca as one way of weakening Germany before invasion rather than as an independent offensive in its own right, the same role that the German Air Force had had before the aborted Operation Sea Lion, the German plan to invade Britain in autumn 1940. Bombing survived as an option not because it was central to the strategic outlook of the Western Allies, but because it was secondary.

The Casablanca Directive

Straightforward as the final decision for CCS 166 has seemed to later historians, the conference highlighted many of the conflicts and arguments that surrounded the bombing campaign in the last months of 1942. The commitment of the two war leaders and their military staffs to a sustained bombing campaign was not a foregone conclusion. Churchill had shown increasing impatience with Bomber Command since August 1942, when he had assured Stalin on his visit to Moscow that a heavy raid on Berlin was imminent. Harris refused to attack the German capital until he had an adequate force at his disposal, and the eventual raid, code-named Tannenberg, took place only on January 17, in the middle of the conference, long after Stalin too had lost patience with the constant delays.⁶ In the end Churchill had to be content with sending Stalin a list of the sixteen German cities that had been attacked between July and September 1942. Not until March 1943, more than six months after Churchill’s first promise, did Stalin finally acknowledge the news that Berlin had been raided.⁷ There was impatience, too, in both London and Washington, with the slow progress of the Eighth Air Force. Churchill thought that American bombers should be allocated to the war at sea and support for the landings of Operation Torch, and that the Eighth Air Force should abandon plans for a daylight offensive against Germany. In December, Spaatz warned Eaker that the Eighth had to start operations “projected into Germany” or face the prospect of diversion to the Mediterranean theater, but the first American raid on a German target was launched only on January 27, 1943, when fifty-nine bombers attacked the port at Wilhelmshaven three days after the end of the conference.⁸ Arnold later reported to his chief of staff that at Casablanca he had been put permanently on the defensive by the British and American delegations “for not having our heavy bombers bombard Germany.”⁹

The air forces’ case at Casablanca had to be made to a disillusioned audience and it had to be made as far as possible in concert. Yet from the autumn of 1942 there were evident strains in the relationship between the RAF and the U.S. Army Air Forces, despite the public commitment to combined operations. Both air forces realized that bombing had to be presented as a more coherent strategic option than it had offered for much of 1942. In September 1942, Portal and Slessor drew up a paper on “Future Strategy,” which argued for a combined offensive that would create the conditions for an easier invasion of continental Europe by weakening German resistance and might knock Italy out of the war entirely, but no effort was made to articulate what kind of bombing was needed and against which targets.¹⁰ Arnold bemoaned the absence of any definite plan from a British air force “without strength in any one place to win decisively.”¹¹ On September 19, 1942, his planning staff in Washington produced a detailed operational plan, AWPD-42, which resembled the British commitment to wearing Germany down prior to a land invasion, but spelled out minutely how this was to be achieved. The American plan committed their bomber force to bomb by day a list of 177 targets vital to the German war effort, dropping 132,090 tons of bombs on 66,045 operational sorties; the seven chosen target systems were the German Air Force, submarine building, communications, electric power, oil, alumina, and synthetic rubber. A counterforce strategy against the German fighter fleet was described as a key intermediate aim, whose achievement would make it possible to complete the rest of the program in time for invasion, but counterforce strategy never appealed to the RAF.¹² Portal told Arnold politely that he had read AWPD-42 “with great interest,” but it does not seem to have brought the two sides to a common view except that bombing mattered to ultimate victory. Arnold complained to the American Joint Chiefs at Casablanca that the British in his view seemed incapable of thinking in global strategic terms but simply chased “the next operation”; he did not think the British “had ever had a definite bombing program,” and at his insistence the Combined Chiefs of Staff were asked to draft a priority bombing program, which provided the trigger for the Casablanca Directive a few days later. Even this amounted to a compromise between the British statement of general aims about undermining German morale and American articulation of a list of priority targets.¹³

The most awkward issue at Casablanca was the argument over daylight bombing. This had been a running sore through 1942 as the Eighth Air Force built up its capability. Churchill was strongly skeptical of the claim that daylight bombing would work. Neither the RAF nor the German Air Force had been able to sustain daylight operations against effective fighter and antiaircraft defenses, and until January 1943 the Eighth Air Force had only flown against light resistance in France. Churchill began a sustained campaign in the autumn of 1942 to persuade the American side that day bombing was too risky over Germany: “They will probably experience a heavy disaster,” he minuted to Portal, “as soon as they do.”¹⁴ Churchill thought it more sensible for American bombers and crew to be converted for night work and integrated with Bomber Command. In October he asked Dwight D. Eisenhower, recently appointed supreme commander for Operation Torch, if American bombers could not be changed over to night fighting.¹⁵ On the advice of both Portal and Sir Archibald Sinclair, the air minister, Churchill restrained himself from pressing the point too far in case the American leadership decided to switch their bombing effort to another theater. RAF leaders waited to see what would happen with daylight attacks before deciding whether to insist that the American air force accept the alternative of night bombing.¹⁶ The American air force delegates at Casablanca knew that this was an argument they had to win. After clearing the request with Eisenhower on January 13, Arnold invited Eaker to fly to Morocco to help him present the case for the American offensive. Arnold warned him at once that Churchill had already suggested to Roosevelt that the Eighth Air Force switch to night bombing under RAF control. Eaker was asked to draft notes for “The Case for Day Bombing”; he prepared a one-page synopsis with seven principal arguments to show to Churchill and a fuller version to help Arnold influence the Combined Chiefs of Staff.¹⁷

Although it was unlikely that Churchill would get his way, given the weight of British and American opinion in favor of trying out the day-bombing experiment, the risk existed that Roosevelt would be too preoccupied with other issues to notice. At a high-level discussion with the president on January 18, Eisenhower and Spaatz secured agreement that neither bomber force should have the right to “alter the technique or method of operating” of the other. American fears that Harris might be placed in overall command of a joint bomber offensive were set aside by the decision to make Portal, who was a popular choice with the Americans, the nominal director of the whole bombing campaign.¹⁸ On January 20, Eaker was given a brief appointment to see Churchill so that he could present his paper. Churchill greeted him dressed up in the uniform of an RAF air commodore and the two men sat down on a couch to talk. The prime minister read aloud the page-long list of reasons for day bombing. Eaker later recalled that when he came to the sentence about round-the-clock bombing, Churchill “rolled the words off his tongue as if they were tasty morsels.”¹⁹ Later that day Churchill was heard to remark, “Eaker almost convinced me,” but he had nonetheless agreed to give day bombing over Germany a preliminary trial. At a meeting that evening Roosevelt and the army chief of staff, General George Marshall, also gave daylight bombing their blessing, and the following day Slessor was able to draft his directive for bombing by day and by night, one of the few features of the subsequent combined campaign on which both sides were agreed.²⁰

The records of the many discussions held at Casablanca give little hint of the arguments over bombing taking place in the wings. In the minutes of the American Joint Chiefs of Staff the bombing campaign was mentioned briefly only three times; during the plenary sessions bombing was discussed on only two occasions, again at no length. In the list of priorities finally agreed to by the Combined Chiefs the critical issues were the commitment to an invasion of Italian territory in the Mediterranean and an eventual campaign in northwest Europe, for which bombing would be a necessary prelude to maximize the chances of success for a major combined-arms operation. Churchill telegraphed the War Cabinet from Morocco the results of the conference, but included no mention of bombing.²¹ The army and navy commanders at Casablanca devoted almost nothing in their memoirs to the arguments over bomber strategy. The projection of airpower against Germany was essentially subsidiary to the wider strategic intention of reoccupying Europe during 1943 and 1944. The Casablanca Directive itself was a brief set of instructions to destroy and dislocate the German “military, industrial and economic system” and to undermine the morale of the German people to the point where the German power of resistance was “fatally weakened.” This was worded in terms that were permissive rather than prescriptive, and its force was immediately compromised by the list of other tasks bombing could be called upon to perform: bombing the submarine bases in France; attacking Berlin to keep the Russians happy; a campaign against Italy when required; objectives of fleeting importance (including German naval vessels); and full support “whenever Allied Armies re-enter the Continent.”²² This was a wish list that encouraged the continued dispersion of Allied bomber forces.

When Eaker arrived back in England on January 26 he ordered the first American raid on Germany for the following day, and then dined that night with Harris to discuss what had happened. Two days later Norman Bottomley, the deputy chief of staff, was asked to send the new directive to Bomber Command. His original letter included the decision to make Portal responsible for the strategic direction of the bomber offensive, but Portal thought it more prudent not to advertise the change to his prickly subordinate, and on February 4, Harris was sent only the Casablanca Directive.²³ Although it was suggested that the new directive replaced the one issued to Bomber Command in February 1942, it is clear that Harris did not regard it as anything more than a statement of intent. In his memoirs Slessor described the Casablanca Directive, which he had drafted, as a policy statement rather than a proper directive.²⁴ Both air forces could read into it what they wanted.

A Combined Offensive? January–July 1943

In the last months of 1942 the term “combined offensive” began to be used more commonly. In August 1942 the Joint Planning Staff had drawn up recommendations for a concerted Anglo-American program of bombing that provided the background for the eventual directive at Casablanca.²⁵The preamble to AWPD-42 stipulated that the offensive was “a combined effort” of the two air forces, the Eighth Air Force concentrating on destroying precision objectives by day, the RAF on night bombing of areas to break down morale. The passage was underlined to give it added force.²⁶The combination was little more than a marriage of convenience. American air forces based their planning and preparation on isolating and destroying Germany’s key industrial and economic targets and eliminating German airpower—much as the German Air Force had done against Britain—while Bomber Command continued, when able, its unremitting destruction of the central areas of German industrial cities.

As in any marriage of convenience, the partners had separate beds. There had been suggestions before Casablanca that there should be a single commander for the bomber offensive. Arnold wanted a supreme air commander for the whole European theater, but the British preferred separate commands in Britain and the Mediterranean, and Arnold waited for almost a year before appointing Spaatz as supreme commander of all American strategic and tactical air forces in Europe in the face of British objections.²⁷ The decision to accept bombing by day and by night underlined the need for two separate organizations, and although Portal had been given overall responsibility for coordinating the bomber offensive, he was not in command of either the Eighth Air Force or Bomber Command. This produced an awkward structure in which it remained unclear exactly the limits of Portal’s power or the degree of collaboration between the two Allied bomber forces. Eaker had made it evident well before the Casablanca Conference that he did not regard the Eighth Air Force as in any sense under British command, though he did submit plans to Portal for approval and looked to him for protection from the demands of other theaters and services. “We always feel,” Eaker wrote to Portal in late August 1943, “that our guardian and greatest friend is away when you are absent.”²⁸ The American air forces in Britain found the formal command lines all the way back to Washington difficult to operate smoothly; in turn, air force officials in the U.S. capital were often poorly informed about conditions in Europe, and frustrated by the long distances. Eaker relied on Portal to supply bases and equipment and benefited from the chief of staff’s familiarity with the offensive and with the political arguments that surrounded it. Harris had none of these difficulties. He communicated regularly with Portal and Churchill in defense of his command prerogatives and tolerated as little interference as possible. The two air forces maintained liaison staff at each other’s headquarters, and on occasion collaborated on a common target, but there was no mechanism for shared command. The American mission statement for the offensive described the bombing as “a joint assignment, completely complementary,” which it was, but it remained combined in name rather than fact.²⁹

The divide between the two air forces was explicit on the question of their strategic priorities. The Casablanca Directive required little adjustment for Bomber Command, which had been attacking German morale and destroying industrial cities for several years with the aim of fatally weakening the enemy. Harris remained doggedly resistant to the idea that specific target systems were strategically valuable in themselves and was hostile to the diversion of his force for other purposes. Bomber Command was committed to accruing a growing register of destruction in German cities in the hope that the attrition might at some unspecified point and in an indefinable way weaken to the point of collapse German capacity to wage war. Harris, unlike the American air force commanders, remained convinced that bombing, combined with Soviet pressure, would bring victory in 1944 without the need for an expensive ground invasion.³⁰

In the discussions following Casablanca, Harris summed up for the American side his strategic achievements so far and his plans for 1943. Essen, he claimed, was “smashed out of recognition” and was out of action for two months; Berlin, Nuremberg, Munich, Cologne, and Wilhelmshaven were “badly knocked about” but not devastated; Hamburg, Duisburg, and Stuttgart had had “lucky escapes.” His plan was to devastate one city and to damage three others badly each month up to September:³¹

This will mean 6 cities “Essenised” and another 18 badly knocked about. Taking cities more or less at random from last year’s chief targets, this might work out as follows:–

Devastated Hamburg (counted as two by virtue of its size), Bremen, Duisburg, Wilhelmshaven, Kiel.

Badly Hit Berlin, Bochum, Cassell [sic], Munich, Nuremberg, Dusseldorf, Cologne, Leipsig [sic], Hanover, Stuttgart, Gelsenkirchen, Brunswick, Emden, Frankfurt, Mannheim, Magdeburg, Dortmund, Essen.

This was a crude strategy, crudely expressed, and it greatly exaggerated what Bomber Command could actually do to a city, including Essen. The ambition was not random only in the sense that most German cities housed some industry and could therefore be subject to attack; it could also be given a scientific gloss by the calculations supplied by the RE8 division and the Ministry of Economic Warfare (MEW) on the estimated economic damage already done and the potential economic gains from further destruction. This had allowed Portal in November 1942 to present the chiefs of staff with the grisly prediction that Bomber Command in eighteen months could kill 900,000 Germans, seriously injure another million, destroy 6 million homes, and dehouse 25 million people.³² The MEW drew up a detailed list of all the important industrial and commercial targets in the so-called Bombers’ Baedeker (after the famous German tourist guides). Each installation was awarded a point score: 1+ for factories of leading importance to the war effort; 1 for major plants in major industries; 2 for minor plants in major industries or major plants in minor industries; and 3 for factories of small importance.³³ These scores were then calculated with population size to produce a league table of German cities that Harris kept with him at his headquarters. The list eventually reached over 100 cities, with “key-point ratings” attached to each one—ranging from Berlin at number 1 with 545 to Wittenberg at number 104 with a rating of just 9. Harris crossed out each city on the list as they were attacked.³⁴

American commanders rejected the idea of city bombing and were skeptical of the claim that morale attacks would diminish the German war effort or create a widespread crisis. Portal tried to persuade Eaker in February 1943 that round-the-clock attacks on cities would be strategically valuable—“heavy blows delivered on German cities have far greater effect than they did”—but Eaker would not be drawn in.³⁵ The gulf between the two strategic conceptions, as has often been emphasized, reflected two very different military cultures. American strategic practice was much closer to the German model than it was to the British. Eighth Air Force officers could be genuinely puzzled by exactly what the British strategic aim was. At a meeting called in the Air Ministry in March 1943 by Sydney Bufton, now promoted to director of bomber operations, the American representative asked for an explanation of what Bomber Command was trying to do: “Was it to kill Germans; to cause them to expend man hours; or was it to do specific damage to some certain installation?”³⁶ There was an awkward pause until Bufton announced that it was to neutralize German man-hours, a strategic commitment not expressed in any directive. When Eaker sent a draft of the American plans for fulfilling Casablanca to Spaatz in April, Bomber Command’s effort was described no fewer than four times as little more than “concentrated attacks against related areas and cities.”³⁷ The assumption in American planning was that Bomber Command would now help the Eighth Air Force, not the other way around.

Unlike Harris, Arnold and Eaker wanted to build on the Casablanca Directive to produce a strategic directive that made greater sense. The driving force behind American planning was the belief that whatever the air forces did over the coming year, they should contribute to making the invasion of Europe possible. The air force mission statement defined “fatally weakened” as “so weakened as to permit initiation of final combined operations on the Continent,” which gave bombing a defined strategic purpose.³⁸ Arnold asked the Committee of Operations Analysts in Washington, set up in December 1942 under the prominent lawyer Elihu Root Jr., to draft a list of targets whose destruction would contribute to Axis defeat; staffed by men from business and the professions, the committee supplied the data on nineteen industrial target systems by March 1943.³⁹ Arnold instructed Eaker to work out with Portal the precise number of targets and the degree of operational effort required to destroy them, and by early April the draft plan for a Combined Bomber Offensive (CBO) was ready. The final version identified seventy-six key targets, with the aircraft industry at the top, submarines second, and ball bearings third.⁴⁰ It was sent on to Bufton by an Air Ministry colleague with a covering note that it was “dull as hell” and needed to be shorter and brighter. It had been drafted chiefly by Eaker and his staff and was based on their fundamental realization that the German air forces had to be defeated first before the subsequent precision bombing of key industrial targets could be carried out without insupportable losses, and it was this argument that separated American from British bombing strategy most clearly. The adoption of a counterforce priority proved to be strategically well founded, as it had been for the German Air Force in 1940.⁴¹ The “intermediate target” became in effect the primary target.

The Air Ministry planning department had already tried to alert Portal to the growing threat of the German single-engine fighter force before Eaker’s report was available, so that there was support from both air forces for the shift in priority. Harris sent Eaker a fulsome response, since the new plan was clearly designed for the American market and impinged little on what he hoped to do.⁴² The report was sent to Washington for Arnold’s approval, and in late April Eaker traveled back to the United States to defend his plan before the Joint Chiefs, who approved it on April 29. This was in some ways a more important moment than the argument at Casablanca, since the whole offensive critically relied on political approval in Washington for speeding up bomber allocation for the Eighth Air Force. Late in May, Eaker was informed from air force headquarters in Washington that the plan had been approved by the president and by Churchill; the Combined Chiefs of Staff endorsed it with Portal’s strong support on May 18.⁴³ The new directive for what was code-named Operation Pointblank was issued to Harris and Eaker on June 10, 1943. The close link with the planning for Operation Overlord, the planned invasion of northern France, was explicit. The Combined Chiefs asked for regular reports on the progress of the CBO to help them judge when conditions for invasion were ripe. Portal set up a regular flow of monthly reports from the Joint Intelligence Committee, while Eaker promised Washington that Arnold would get an analysis every two weeks together with a monthly summary. In September the Combined Chiefs confirmed that the CBO was now the “prerequisite to ‘Overlord,’” with the highest strategic priority.⁴⁴Air supremacy was the key to successful invasion, and bombing was its instrument. In America there was at last a sense that a proper air strategy was in place. Assistant Secretary of War Robert Lovett wrote to Eaker in July that the combined offensive “ought to have the effect of the famous old ‘one-two’ in prize fighting.”⁴⁵

These differences in command and strategy were compounded by the gulf separating the two forces in terms of current striking power. Bomber Command had been slowly expanded and modernized for three years before Casablanca and was closer to being able to redeem some of its tarnished promise in the spring of 1943 than the Eighth Air Force, which was still in the difficult throes of constructing a viable organization. Building up an effective bomber force differed markedly from the development of a major land army. Bomber commands consisted of volunteers with a high degree of long-term, expensive, and specialized training. The initial equipment was complex and industrially demanding. On both counts loss rates had to be kept as far as possible to a supportable minimum. The nature of air battle required numerous well-equipped permanent bases, an extensive maintenance organization, a large stock of spares, and, in the case of the Eighth Air Force, a long transoceanic logistics tail. The sharp end of air combat was supported by a ground organization many times larger than the aircrew on which the campaign rested. There was no supporting infantry in air combat.

Bomber Command became a substantial force only in the spring of 1943 as the changeover from medium to heavy bomber production was finally completed. By March 1943 it was planned to have forty-nine heavy-bomber squadrons, by June as many as sixty out of a worldwide total of 431 RAF units of all types. Heavy-bomber squadrons made up 7 percent of RAF squadron strength in September 1942, 14 percent by the summer of 1943.⁴⁶ This was still far short of what Harris had wanted. In January 1943 there were still only an average of 514 heavy bombers and crews operationally ready. Harris’s command reached its peak strength in heavy bombers only in 1944 and 1945; the same was true for pilot strength, which was not much greater in 1943 than it had been in late 1941, though each of the heavier aircraft they flew dropped up to four times the weight of bombs (see table 2.1 for the growth of Bomber Command). The overall size of the command was dictated by the need for specialized ground personnel, which meant by the start of 1943 a combat strength of 23,000 aircrew (including training and OTU personnel) and a supporting force of 138,000 men and women, a ratio of 1:6. As the force of heavy bombers and pilots expanded, the ratio changed. By the end of the war there were 49,000 aircrew supported by 174,000 ground staff, a ratio of 1:3.5.⁴⁷ Women made up 17 percent of the force by 1944.

Table 2.1: Bomber Command Strength in the United Kingdom, 1939–45

Figures for columns 1–3: January 1 each year. Figures for columns 4–7: July 1940, December 1941–44, May 1945. Column 8: February 1942, January 1943–45.

Source: Compiled from TNA, AIR 22/203, War Room Manual of Bomber Command Operations, 1939–1945, chart 1; AIR 20/2025, Air Ministry Statistics, RAF personnel, establishment and casualties, 1939–45; UEA, Zuckerman archive, SZ/BBSU/3, Exercise Thunderbolt, précis no. 10, “Administrative Aspects of the Bomber Offensive.”

The aggregate figures disguised manpower problems. By the summer of 1943 there were substantial shortages of skilled labor, much of it required by industry or by overseas air squadrons. This included a deficiency of 65 percent of aircraft fitters (I Class), the most important category, and an overall shortage of two-fifths of skilled technicians and more than a third of all other trades.⁴⁸ There was also a persistent shortage of construction labor for airfields, for both Bomber Command and the Eighth Air Force. Heavy bombers required larger airfields, solid runways, and extensive depots. In January 1943 there were 32,000 workers building airbases for the American force, 42,000 for Bomber Command. As demands for training facilities also expanded, more new bases had to be built. In addition to the 81 Bomber Command operational fields ready by January 1944, there were also 47 for training purposes.⁴⁹ The Eighth Air Force initially calculated that it would need 61 completed airfields by the end of 1943, but eventually built 120, utilizing 1 million man-months of labor and laying 46 million square yards of concrete. The Royal Canadian Air Force No. 6 Group was promised 15 airfields, but ended up with only 10.⁵⁰

The pressure on airfield space was reduced by the decision very early in the war to disperse most of the basic training overseas. It is seldom sufficiently acknowledged that the British bombing effort during the war was in reality a British Commonwealth undertaking. Britain was never “alone” during the Second World War. On December 17, 1939, an agreement had been signed with the Canadian government to set up the British Commonwealth Air Training Scheme on bases in Canada. During the course of the war a peak of seventy-three schools were set up under the scheme, with a further twenty-four under RAF control. The scheme turned out 131,000 trained aircrew, including 49,808 pilots and 29,963 navigators. By 1944 over 3,000 completed training each month. The majority (55 percent) went into the Royal Canadian Air Force (a high proportion for Canadian units in Bomber Command), while the RAF took one-third; the remainder went to the Australian and New Zealand air forces.⁵¹ Other British aircrew were also posted to training schools in the United States under the so-called Arnold Scheme. In April 1941 the American army agreed to allow British participation in the Southeast Air Corps Training Program in the southern United States, and 7,885 pilots were sent for basic training, together with 1,200 navigators. The program was linked with Canadian training, but the failure rate in the United States was high because of the poor level of scientific education among British recruits, and the scheme petered out in 1942.⁵² Although crew were trained for a variety of different combat roles, the most pressing need was for bomber crews, and it was here that overseas training had its greatest impact.

The training program was the prelude to a great expansion of Commonwealth and European participation in the bombing war. There were Free French, Dutch, Norwegian, Czech, and Polish crews in Bomber Command, but by far the largest contingents came from the main Dominion states: Canada, Australia, and New Zealand. Only the Canadian units were large enough to be organized into a separate national organization, because Canada, unlike Australia and New Zealand, was not directly menaced by Axis aggression and could confidently send its forces overseas. The RAF had not initially warmed to the idea of separate national units, partly because the complex training pattern made it difficult to keep crews from the same nationalities together as they went through the system. This was the problem with the limited number of Australians who contributed to the bombing war from British bases. Three Australian squadrons were activated, but even these could not be replenished regularly with only Australian airmen, while two-thirds of the ground personnel were British. In the spring of 1943 an effort was made by the Australian War Cabinet to stimulate creation of a distinctly Australian bomber group; there was resistance from the crews, already integrated with RAF units, and eventually an acknowledgment that the personnel needed to operate a full group could not be freed from the limited labor supply in Australia needed for the Pacific War and wartime industry.⁵³ The pressure from the Canadian government for the “Canadianization” of the units organized in Bomber Command had more success. On January 1, 1943, an entirely Canadian group, No. 6, was activated under the command of Air Vice Marshal G. E. Brookes. It was spread out across the Yorkshire countryside, with headquarters at Allerton Hall, nicknamed “Dismal Castle” on account of its gloomy aspect. In total, fifteen Canadian squadrons were formed, though like the Australian units they were not composed entirely of Canadians. The exception was the French Canadian squadron, formed in autumn 1942, where a great effort was made to ensure that it received only French-speaking aircrew.⁵⁴

The Eighth Air Force was, by contrast, an almost entirely American effort. The structure of the force set up by Spaatz and then Eaker differed from Bomber Command because it also included its own fighter, training, and air service commands, which were regarded as integral to the bombing campaign. When Eaker replaced Spaatz as commander of the Eighth Air Force late in 1942, he appointed Colonel Newton Longfellow, commander of the Second Bombardment Wing, to take over the Eighth Bomber Command. The force was divided into three air divisions and then into combat wings, each with its own tactical headquarters, each wing made up of three heavy bombardment groups, or squadrons, and modeled, despite the differences in vocabulary, on Bomber Command practice. The force was still very small at the start of 1943, partly because of the decision to take a large component of aircraft and crew to establish the Twelfth Air Force in the Mediterranean theater. Some 27,000 men and 1,072 aircraft were transferred at a critical point in the buildup of the offensive, leaving Eaker temporarily with just 27,000 men and 248 heavy bombers.⁵⁵ The diversion left the Eighth Air Force as little more than a skeleton. By April there were still only 250 heavy bombers, of which around half were serviceable at any one time, a reflection of the difficulty in establishing an effective supply organization and the need for extensive modification of the B-17s for actual combat conditions. Even by June 1943 the serviceability rates for heavy bomber units was little more than half.⁵⁶ In the second half of 1943, however, the Eighth Air Force began to expand rapidly, reaching its peak strength, like Bomber Command, in late 1944 and early 1945; see table 2.2 for the buildup of the Eighth Air Force.

Table 2.2: The Buildup of the Eighth Air Force, 1943–45

*Figure for December 1943.

Source: AFHRA, Maxwell, AL, 520.056-188, Statistical Summary of Eighth Air Force Operations, 1942–1945; CD A5835, “Eighth Air Force: Growth, Development and Operations 1 December 1942–31 December 1943,” Personnel Status, Exhibit 1.

Even more than the RAF, the American organization relied on very extensive base facilities and a large noncombat cohort to service and maintain the force. Among the many problems facing the Eighth Air Force in the first half of 1943, the issue of supply—everything from aircrew to spare parts—was the most pressing. For Bomber Command the logistics were straightforward. For the Eighth Air Force all the combat matériel, except for heavy aircraft, which could be flown across the northern Atlantic, had to come by ship and be stored in large service depots set up at each bomber base to handle the incoming resources. A truck transport system was established that by the end of 1943 could move 1.5 million ton-miles each month. The seven principal storage depots covered an area of more than 9 million square feet.⁵⁷ The most urgent need was for service personnel. In the late spring Arnold sent Major General Follett Bradley, inspector general of the Army Air Forces, to England to help organize an effective program to supply the trained manpower necessary to keep the Eighth Air Force flying. The “Bradley Plan” called for 190,000 personnel in the Service Command to match the planned size of the force, but the figures were always behind target in 1943, partly because around 88 percent of the air force had to be shipped across the Atlantic in competition with vital supplies for the Mediterranean and the buildup of forces for the eventual invasion of France.⁵⁸ By June 1943 five bomber groups had arrived in England without their ground crews or operational equipment and had to double up on bases that did have them.⁵⁹ The Service Command rejected the offer of help from the RAF and insisted that American aircraft should only be maintained by American workmen, but many aircraft arrived in England in need of extensive modification. To Arnold’s complaints about the low level of combat readiness of the Eighth Air Force by the summer of 1943, Eaker retorted that aircraft should have been prepared for combat in the United States rather than rely on modification depots in England, which lacked the means to convert aircraft quickly.⁶⁰ It took most of the year before serviceability and replacement rates for American heavy bombers could keep more than half the air force flying.

The Eighth Air Force also suffered from the absence of a large corps of trained officers and men from the prewar air force. Some senior airmen had combat experience from the First World War, but most younger officers had never had to fire a gun or drop a bomb in anger. Arnold acknowledged that with only 1,500 regular officers in 1941, spread worldwide a year later, “the experience level is very thin.”⁶¹ Tactical thinking about the conduct of long-range bomber operations was in its infancy and relied on learning from both the British and German experience. The large number of volunteers in the United States for service in the air force had not, unlike those in the RAF, witnessed the battles over northern France and southern England or seen what bombing might achieve. Eaker called them “sturdy amateurs,” not lacking in enthusiasm or courage, but not the equivalent of a highly trained peacetime force.⁶² Brigadier General Haywood Hansell, temporarily in command of the Eighth Bomber Command before Longfellow’s appointment (hitherto commander of the First Bombardment Wing), produced a report for Eaker in February 1943 suggesting that the force was simply not ready yet for a major offensive against Germany. Hansell added that for the following months crews should be asked to undertake shallow attacks against German targets to keep losses to a minimum and to practice with the new target-finding apparatus, Oboe and H2S, taken over from the RAF.⁶³ Nevertheless, the pressure on the command to show that it could produce results made it difficult for Eaker to limit what was done, even when the early raids sustained high losses and resulted in extensive combat fatigue. Between January and April the monthly loss rate averaged almost 7 percent, too high a figure for such a small force.⁶⁴ The slow expansion of the bomber units meant that crews arriving together from the United States might be broken up to fill the losses in existing groups. This process, Eaker told Arnold, created “an understandable and definite loss of morale.”⁶⁵

The problems faced by the Eighth Air Force in adjusting to new conditions of combat, thousands of miles from the United States with thousands of freshmen aircrew, were exacerbated by the often poor relations between American airmen and the British communities that had to accommodate and service them. The relationship with the air force suffered from the instinctive sense of competition between the Eighth and Bomber Command, since they were the only servicemen actively in battle against the German homeland. The numbers involved were also very large—by December 1943 there were 283,000 servicemen and civilians in the Eighth Air Force—while the irregular nature of air combat gave aircrew long periods when they lived among predominantly civilian communities. One woman, keeping a diary for the British Mass Observation public opinion surveys, condemned Americans as “loud, bombastic, bragging, self-righteous.” Opinion polls showed that the British public most disliked American “boastfulness,” “immaturity,” and “materialism.”⁶⁶ Though established prejudices might explain some of this reaction, the behavior of mostly very young men a long way from home, learning “British English,” contributed to the confrontation. Eighth Air Force commanders were warned in December 1942 that “unbridled speech” was causing embarrassment to Anglo-American relations and that all supplies of alcohol to the command would be suspended if it continued.⁶⁷ General Marshall wrote to all senior commanders that American officers had encouraged a “marked hostility and contempt for the British,” and great efforts were made to introduce activities that would educate the American servicemen into treating their British hosts with greater respect and dignity. The provost marshal of the Eighth Air Force, in a lengthy report on the conduct of airmen toward the British, complained of the persistent “Limey complex” that made them largely indifferent to their hosts unless there was the prospect of sex. A Special Service Study found that only 2 percent of the American airmen had actually visited a British home.⁶⁸ The Eighth Air Force set up a speakers’ pool with officers assigned to give talks to British audiences about American customs and manners; between April and September 1943 the speakers averaged fifteen lectures a month, a grand total of 241. Eaker joked that out of the three possible crimes his men might indulge in—murder, rape, and “interference with Anglo-American relations”—the first two might under certain circumstances be pardoned, “but the third one, never.”⁶⁹

British criticism was not confined to popular areas of social friction. Over the first half of 1943 there was evidence of mounting frustration from the Air Ministry and Bomber Command over the long apprenticeship and unredeemed promises from the American air force. Harris and Portal both pressed Eaker to collaborate more fully in the renewed offensive in the spring of 1943, though both well knew that Bomber Command had taken three years to reach its current size and capability. There were British complaints about the quality of the B-17 Flying Fortress as a daylight bomber and arguments over the effectiveness of the new American oil-based incendiary bomb, the six-pound M-69. British tests suggested that the new bomb lacked penetration, had an unreliable fuse, and suffered from relatively low incendiary efficiency, and despite American objections to the test results, the RAF continued to use the standard four-pound incendiary.⁷⁰ There were strong objections to the British view of the B-17, but the American judgment on the bombers currently employed also accepted that the two main bombers, the Flying Fortress and the B-24 Liberator, were rapidly reaching the end of their useful life.⁷¹ Indeed, the harshest criticism of Eighth Air Force performance came from the American side, not from the RAF. Arnold, whose health deteriorated badly during 1943, finally lost patience with Eaker’s force in June 1943. Citing the number of aircraft already sent to Britain, Arnold telegraphed Eaker that the number of bombers sent out on each mission “is not rpt not satisfactory,” and told him bluntly to change his commanders and staff to find officers who could organize attacks in “a highly efficient manner.”⁷²

In the short correspondence that followed, Eaker was defensive about his force and resentful of Arnold’s intervention. Arnold ought certainly to have known better, since the ratio between aircraft supplied and aircraft on missions always had to take account of those awaiting repair, modification, or use for training purposes. “Neither of us,” Eaker replied, “has been able to accomplish ideal for reasons both should appreciate. We get nowhere with recrimination.”⁷³ Nevertheless, Eaker agreed to make major changes in his command structure, anxious perhaps about the security of his own position. Longfellow, whom Arnold regarded as insufficiently aggressive, was redeployed to Washington; Hansell was passed over (too “nervous and highly strung,” according to Eaker); responsibility for the Eighth Bomber Command was given on July 1, 1943, to Brigadier General Frederick Anderson, commander of the Third Bombardment Wing, while Hansell was replaced as commander of the First Bombardment Wing in June. The Third Wing was taken over by a young commander admired by both Arnold and Eaker, Colonel Curtis LeMay, future chief of the Strategic Air Command after the war. Anderson had arrived in England in February, flew his first mission to Rouen on March 15 (“enjoyed it thoroughly”), and in his diary looked forward to the day when a stream of B-17s would be taking off every day “as they carry out their joy, dropping bombs on Germany.” He flew regularly with his crews and took the same risks they did.⁷⁴ Though Anderson had only been a few months in active service, Eaker judged that he had the right character for high command, though he regretted losing Longfellow, who was a close friend. “This Bomber Command job of ours is a killer,” he told Arnold. “It will break anybody down in six months unless he is a very unusual fellow.”⁷⁵ What had most concerned Arnold during the command crisis was not so much the virtue of the bomber offensive itself and its achievements as the extent to which it represented the claims to a distinct air force identity and a distinct strategy. For Arnold this meant a clear demonstration to the American public that the offensive was producing results that would inspire their “faith in our way of making war.”⁷⁶

The different force profiles and levels of preparation were no more sharply evident than in the contrast between the operational performance of the two bomber forces over the course of 1943, when Harris launched what have come to be seen as the three major battles of the bomber offensive, against the Ruhr-Rhineland in the late spring and early summer of 1943, against Hamburg in July, and against Berlin in the autumn. The campaign against the Ruhr-Rhineland was not in any sense a new one, since it had been a first priority for bombing ever since the onset of the campaign in May 1940. The difference in the spring of 1943 was the advent of substantial numbers of heavy bombers, particularly the Lancaster. During the months from March to June, when the Ruhr battle took place, the operational strength of Bomber Command averaged 794 heavy bombers, of which 578 (or 73 percent) were serviceable. Each bomber dropped an average of just over 7,000 pounds of bombs, against an average load of 4,970 for the Eighth Air Force throughout 1943.⁷⁷ The Pathfinder Force was now fully equipped with the Mosquito Mk IX, which utilized the new electronic aids, Oboe Mk IA and H2S, which were both now available in sufficient quantity, though H2S proved a disappointment over heavily urbanized areas. Raids on the Ruhr had always been subject to the hazards of cloud and industrial smog and the effective use of decoy sites. Navigation equipment no longer dependent on visual sighting was expected to allow a much greater degree of concentration at or near the principal aiming point. Over the target, new systems of marking had been developed using bright white markers when the ground was visible, or red, green, and yellow sky markers when it was not. The colorful flares were known by the code name Wanganui; the German public called them “Christmas trees.” Once at the target the new Mk XIV bombsight could be used even when the aircraft took evasive action.⁷⁸ Radio countermeasures were available to block the German Freya radar, code-named Mandrel, and to interfere with German ground-control transmissions, known as Tinsel, while new airborne devices to warn of German night fighters and radar (“Monica” and “Boozer”) were finally being used, though again with only mixed effect.⁷⁹All of these many technical and tactical innovations made Bomber Command a much more formidable threat. In a speech later in the war, Harris admitted to an army audience that the bomber offensive only started seriously in March 1943.⁸⁰

The first major raid on the Ruhr was made on the night of March 5–6 against Essen. A force of 442 aircraft attacked with 1,014 tons of bombs, two-thirds of them incendiaries. The Pathfinders worked well and an estimated 75 percent bombed within three miles of the city center, which was the principal aiming point. The apparent success of the raid gave Harris his verb “to Essenise” the target, but the next raid on Essen was less successful, and a major raid on Duisburg on March 26–27 was scattered, largely because five out of nine of the Pathfinder Mosquito aircraft had to turn back with technical problems. A major raid on Berlin, conducted at Churchill’s suggestion, lost direction and hit the capital with only ten high-explosive bombs; the next raid missed the city altogether. Although the Ruhr plan was the first time Bomber Command had concentrated on a single coherent target, major raids were also made on more distant cities, reducing the impact on the Ruhr-Rhineland to no great effect. Attacks on Munich and Stuttgart recorded between one-fifth and one-third of bombs within three miles of the aiming point, levels of accuracy not significantly better than the year before.⁸¹ Between March and June, when the battle against the Ruhr came to an end, Bomber Command had launched twenty-eight raids against cities in the Ruhr-Rhineland area, but another eighteen raids against targets in central Germany, Italy, and France. Though the pattern of destruction varied widely between the different operations, the attacks on the Ruhr-Rhineland were the first raids to inflict serious levels of destruction on the urban area.⁸²

The reaction from the German side was to search for new ways to cope with the sudden escalation of RAF capability. Hitler was preoccupied with stabilizing the war in Russia and North Africa but angered by what he saw as the continued failures in air defense. His air adjutant Nicolaus von Below recalled long evening conversations with Hitler about the inadequacy of antiaircraft fire, the incompetence of Göring, and the shortage of modern aircraft designs, but Hitler remained, according to von Below, “at a loss” on questions of airpower.⁸³ To stifle widespread popular anxiety, Joseph Goebbels, “Minister for Popular Enlightenment and Propaganda,” ordered the press and propaganda agencies to stop using the word “mood” to describe public attitudes, which could evidently be widely variable, and to write only about “high morale.”⁸⁴ Security Service intelligence reports showed the population in the bombed regions dismayed and restless at the apparent absence of effective defense, while exaggerated rumors of the extent of the destruction and casualties in the Ruhr were in circulation throughout the unbombed areas and could not be stopped. “Even sensible people,” ran one report, “have given these rumors credence.”⁸⁵

The German defenses against night bombing were nevertheless stronger than they had been a year before. General Josef Kammhuber had around 400 night fighters, double the level of the previous year, organized in five wings. There were around 500 day fighters in the Western theater protecting the Reich against daylight incursions. Each box in the Himmelbett system now had enough radars to control three fighters at a time, and had developed the means to pass on information to neighboring boxes, but the rigid nature of the fixed line of air defense made less sense against large concentrations of heavy bombers that could routinely swamp the line as they crossed into Germany. Kammhuber proposed a single central authority to control the whole night-defense system and a fivefold increase in the night-fighter force, but the proposal was rejected by Hitler in favor of strengthening antiaircraft and searchlight defenses around the vulnerable inland areas.⁸⁶ As a result the night-fighter force stagnated, taking losses of 282 aircraft during the period of the Ruhr battle, against an overall loss of 600 RAF bombers from all causes. Pressure to change to a more flexible system of defense was rejected. The idea of using ground-control stations to direct night fighters into the bomber stream so that they could fly with it, shooting down any bomber that came within range, a tactic known as Zahme Sau (“Tame Boar”), was turned down because it would drain resources away from the Kammhuber Line; a second proposal from Major Hans-Joachim Hermann to use single-engine day fighters at night against bombers illuminated by searchlights, flares, and target markers, known as Wilde Sau (“Wild Boar”), arrived too late for the spring attacks and was again difficult to integrate with the existing system.⁸⁷ Kammhuber’s insistence that his aerial fortification was the only way to combat the bombers brought him into conflict with Göring and his deputy, Erhard Milch, and in November he was finally relieved of his command and sent to run the rump Air Fleet 5 in Norway.

In the midst of the campaign against the Ruhr, Harris was compelled to organize an operation of which he fundamentally disapproved. The engineer Barnes Wallis had begun work in 1940 on the kind of explosive device needed to breach a major dam wall. In March 1941 an Air Attack on Dams Committee was set up to study the possibility, using the Road Research Laboratory as a base. In April 1942, Wallis developed the idea of a cylindrical bomb, dropped from a low height, to bounce across the water and drop to the foot of the dam wall; code-named Upkeep, the bomb tests represented a real technical challenge but were convincing enough by February for the air staff to approve a possible operation against the German Möhne, Sorpe, and Eder dams.⁸⁸ When the report was passed to Harris, he scrawled at the bottom, “This is tripe of the wildest description . . . not the smallest chance of it working.”⁸⁹ Nevertheless, more trials were conducted to determine whether the “bouncing bomb” was a viable operational proposition. A squadron of Lancasters under Wing Commander Guy Gibson, no. 617 squadron, was activated on March 21, 1943, and trained rigorously for an operation against the reservoir dams that supplied water for the Ruhr. Harris remained adamantly unconvinced: “As I always thought,” he minuted in April, “the weapon is balmy [sic] . . . get some of these lunatics controlled or if possible locked up.”⁹⁰ His skepticism was ignored. On the night of May 16–17, under the code name Operation Chastise, nineteen Lancasters were dispatched, of which twelve attacked the three dams, breaching the Möhne and the Eder, but doing only superficial damage to the Sorpe. Two-thirds of the water escaped from the reservoirs and 1,294 people were drowned in the inundation, including 493 foreign workers. The destroyed dams lost an estimated 25,000–30,000 tons of masonry, but both were repaired by October, while the long-term damage to the industrial water supply was less than had been hoped.⁹¹ Further attacks were ruled out, partly because of the high loss rate. Only eight Lancasters returned from the raid and 56 out of the 113 aircrew were lost.

The early raids against the Ruhr were regarded by Harris as less than satisfactory. The proportion of aircraft dropping bombs within three miles of the aiming point varied greatly. The experiment with sky marking showed that out of eleven raids that used it, one resulted in severe damage, one in considerable damage, eight in “scattered damage,” and one in no damage at all. The photoreconnaissance evidence on major raids showed that the percentage of hits within three miles varied from 80 percent to 25 percent, with most less than 50 percent.⁹² Harris assumed that “weaker brethren,” as he called them, were failing to press home attacks, while he singled out the tactic of violent evasive action (an instinctive response to intense searchlight and antiaircraft activity) as a key culprit. In early May 1943 he circulated a tactical memorandum to groups on “Evasive Action by Bombers” to replace existing instructions, which had given pilots advice on what forms of evasion to take under different circumstances. The new memorandum insisted that most forms of evasion were “useless,” since an antiaircraft barrage was indiscriminate and inaccurate (a claim scarcely justified by the 1,496 bombers damaged by antiaircraft fire during the campaign). Evasion, the report continued, increased the chance of collision, kept bombers longer in the danger zone, and placed serious strain on the aircraft structure. Crews were told to fly “straight and level” through the target area to increase the concentration of bomb hits.⁹³ Whether crews really did respond to the instruction, or whether the growing experience of the Pathfinders and the surviving pilots reduced losses and increased concentration, the major raids carried out between May and July against Cologne, Duisburg, and Wuppertal-Barmen resulted in the most destructive and lethal attacks so far.

These last raids produced very heavy damage to the centers of the cities, and in the case of Barmen, attacked on the night of May 29–30 by 719 bombers, created a major conflagration that consumed four-fifths of the built-up area and killed 3,400 people, the largest number in any one raid until then. Wuppertal, the other half of the town of Wuppertal-Barmen, was raided on June 24–25 by 630 aircraft and 94 percent of its urban area was destroyed or damaged, with 1,800 deaths. In the four months of the heavy “crash raids,” as they were called, an estimated 22,200 were killed, almost twice the number killed in the whole period since May 1940. Some 55,700 buildings were rendered for the moment uninhabitable. “They had ruined the Ruhr,” complained Hitler in June, though only 5 percent of buildings were actually destroyed.⁹⁴ The exact nature of the damage could not be known to Bomber Command, but intelligence estimates of the damage to the Ruhr were constructed in the summer to give Harris the necessary matériel to defend his city-bombing campaign. This was difficult to do and had to be based on extrapolation from the detailed statistical surveys of housing damage and man-hour losses in British cities bombed in 1941 and 1942. The Joint Intelligence Committee estimated that 9 percent of the population of the most heavily raided cities were homeless (422,500) and 38.5 percent (1,816,000) had housing damage. The committee estimated that 68,750 houses had been destroyed, a figure closer to the reality than might be expected from simple photoreconnaissance. Estimating the impact on the economy was more difficult. The Krupp works in Essen were thought to have lost between one-quarter and one-half of planned output over the summer. The MEW estimated a total loss to German production of 10–12 percent from attacks on the Ruhr-Rhineland (including 2–2.5 million tons of steel) and a bad state of morale.⁹⁵ A study by the RE8 department two months later of the impact of the raids on Essen, based on studies of British workers during the Baedeker raids on historic British cities in the spring of 1942, concluded that if German workers reacted in the same way as British workers to the loss of housing and amenities, then Essen lost the equivalent of fifty city-days of work. None of these efforts at calculating what Bomber Command was achieving were coordinated or consistent. Nor could they confirm whether bombing was a strategically sensible use of British resources.⁹⁶

Alongside the night bombing, the Eighth Air Force, as one senior American commander put it, “can only nibble at the fringes of German strength.”⁹⁷ There were three small raids on German territory in March; one in April (a mission to Bremen that cost the loss of 15 percent of the force); six attacks on the north German coast in May; and three in June, including the only attack on the Ruhr, against the synthetic rubber plant at Hüls on June 22, where the loss rate was almost 9 percent.⁹⁸ The two forces were at quite different stages. In the first six months of 1943, Bomber Command dropped 63,000 tons of bombs, the Eighth Air Force only 8,400 more. For the American command, Operation Pointblank had not yet started; in contrast, the RAF, as the director of plans explained in March 1943, hoped that bombing “may well produce decisive results this year.” There was nevertheless no prospect that the two bomber forces would constitute what the Joint Intelligence Committee called after the Ruhr offensive “an organic whole,” because of continued American resistance to the idea that bombing cities was strategically useful. The Ruhr battle was one of a long line of Bomber Command campaigns in which the Eighth Air Force was largely absent.⁹⁹

Operation Gomorrah: The Destruction of Hamburg

Back in November 1941 the Directorate of Bomber Operations in the Air Ministry had concluded from a study of urban targets in northern Germany that Hamburg was the city that presented the best general conditions for a large-scale incendiary attack. Since the area was large, “saturation point” would be harder to achieve. The report continued that if Hamburg was selected for special incendiary attack, “the whole of the effort should be confined to the congested city and housing areas North of the Elbe.” To get a clear picture of what saturation point might look like, the directorate used a map of central London as a model, placing transparencies with a typical incendiary bomb salvo printed on them over the London streets to work out how widely the bombs would spread and how vulnerable the buildings might be.¹⁰⁰ Over the winter of 1941–42 a great deal of effort went into working out how to create a major conflagration with the existing technology. In February 1942 a further memorandum on choosing a German city to burn down had Hamburg at the top of the list, its vulnerability rated “outstanding.”¹⁰¹

Harris might well have tried to exploit that vulnerability in May 1942 when he planned the first 1,000-bomber raid, but poor weather prevented it. Had he done so, Hamburg would certainly have sustained far less damage and loss of life than proved to be the case when Harris finally decided, on May 27, 1943, that it would suffer next what he was inflicting on the cities of the Ruhr-Rhineland. What resulted in a series of raids appositely titled Operation Gomorrah was the single largest loss of civilian life in one city throughout the whole European war, exceeded only by the 250,000 Japanese killed in the firebombing of Tokyo and the atomic attacks on Hiroshima and Nagasaki. The destruction of Hamburg in an uncontrollable firestorm on the night of July 27–28, 1943, is often presented as if it were an accident, the result of exceptional meteorological conditions and the failure of German defenses, and not a product of deliberate intention. This is to misunderstand entirely the purpose of the city-bombing campaign, which was predicated from the start on causing as much general damage and loss of life as possible by means of large-scale fires. The firebombing of Hamburg was not exceptional. Not for nothing was its vulnerability rated “outstanding”; it was expected to burn well.

To understand the capacity of Bomber Command to inflict a conflagration of such intensity on Hamburg, it is useful to place it in the context of the long operational and scientific effort devoted by the Air Ministry to understanding how incendiarism worked. When the RE8 department was established in the summer of 1942, one of the first tasks given to it by the Air Ministry was research into the spread of fire and the role of wind speeds in turning a strong blaze into a conflagration. It was assumed that city fires could be acted upon much as a pair of bellows on a domestic hearth if the wind speed and direction were favorable. “Once a large conflagration has become established,” wrote one of the department’s scientific advisers, “the ‘fire-storm’ which it induces is sufficient to ensure its further spread.”¹⁰² On the advice of J. D. Bernal, wind trials were conducted by the Porton Down Experimental Station in Wiltshire using models of urban areas, while the RAF Photo-Interpretation Section provided night photographs to help in identifying the factors that accelerated the spread of fire.¹⁰³ It was also necessary to determine scientifically the vulnerability to firebombs of German urban structures, in particular town houses and apartment blocks, and the quantity and mix of bombs necessary to overwhelm the fire services in a single raid. Work on German structures began in 1942 and was more or less complete by the time of the final RE8 report on “German Domestic Architecture” issued early in 1943. Tests on models of different kinds of German roofs were carried out at the Road Research Laboratory at Harmondsworth, outside London, while material on German staircases and stairwells was supplied from a German publication on “Residences and Houses of the Middle Classes.”¹⁰⁴ The study of German architecture had involved a good deal of argument over the average thickness of wood beams and the penetrative power of the standard four-pound incendiary bomb, which had only been resolved by recruiting émigré German architects, including the Bauhaus founder, Walter Gropius, to confirm the details of construction. The conclusion from studies of roof coverings, joists, and floor density typical of construction in northwestern Germany and Berlin was that “a German house will burn well.”¹⁰⁵

The most important issue was to decide what mix and weight of bombs would best achieve a conflagration that was beyond the control of the ground defenses. The quantity of incendiary and high-explosive bombs judged to be necessary to destroy each square mile of the target city was cranked up regularly during the two years before Operation Gomorrah. Initial research suggested that between 100,000 and 200,000 four-pound bombs would swamp any fire-watching and firefighting force, but they had to be dropped in large salvos, not in small packets.¹⁰⁶ By late 1942 incendiary technique was better understood. Its purpose, as one Air Ministry report put it, was “the complete destruction by fire of the built-up area of a city.” This required a fire-raising group to isolate the target and start fires pronounced enough for the follow-up force to drop 25,000 incendiaries for each square mile attacked as well as high-explosive bombs to destroy windows, crater the streets, and intimidate civil defense and fire workers. In order to start a conflagration the target area had to be the most densely packed residential areas of the city center, Zones 1 and 2A on the zone maps supplied to Bomber Command. Igniting the “terraces of box-like buildings dating from the Middle Ages” was, according to the Bomber Directorate, expected to “yield good dividend.”¹⁰⁷

To prevent effective firefighting, the incendiary load had to contain not only regular incendiaries but also the delayed-action explosive incendiary, capable of maiming or killing enemy civil defense workers and deterring them from action. These devices were deliberately timed to detonate at irregular intervals, some after three minutes, a small proportion only after ten.¹⁰⁸ In late 1942 a small antipersonnel high-explosive bomb with a trigger fuse was developed, which could be activated by any object, even a jet of water, and would kill those immediately around it without warning.¹⁰⁹It was suggested that a high proportion of delayed-action bombs should be used in “incendiary attacks on virgin towns” to create a powerful deterrent effect on the enemy emergency services.¹¹⁰ There was also considerable argument about the merits of complementing the conventional magnesium bomb with larger oil-based incendiaries, which were also subjected to rigorous scientific testing. The result was the selection of the thirty-pound Mk II containing a mix of white phosphorus and benzol gel, capable of greater penetration than the four-pound incendiary, and effective in spreading fire quickly over a wider area.¹¹¹ All of these bombs were eventually produced and used in substantial quantities over the last three years of the war, peaking in 1943 (see table 2.3).

Table 2.3: Incendiary Bombs Dropped by the RAF, 1940–45

Source: TNA, AIR 22/203, War Room Manual of Bomber Command Operations, 1939–1945, 54.

There was strong American interest in the development of incendiary bombing. The conventional view that the Eighth Air Force used predominantly high-explosive bombs for precision air attacks in contrast to the fire-raising tactics of Bomber Command is not borne out by the evidence. The Bomber Directorate in the Air Ministry collaborated closely with the American Office of Scientific Research and Development, which produced studies on the “Theory and Practice of Incendiary Bombing” and supplied the British with illustrated volumes on major conflagrations in American cities, which were more common and more destructive than in Europe. The photographs resemble closely the aftermath of the major city bombing in Germany.¹¹² In December 1942, America’s foremost expert on fire in foreign countries, Boris Laiming, filed a report that was passed on to the Air Ministry in December 1942. He argued that the only way to start a major conflagration under conditions normally prevailing in Germany was to start a great many smaller fires simultaneously on long strips of urban territory on a day when there was a reasonable wind and low humidity.¹¹³In the United States, experiments were made at the Chemical Warfare Service depot in Utah in burning down different kinds of structures and at a facility in New Jersey on penetrating simulated German roofing with incendiaries.¹¹⁴ American expertise was also invited to Britain. Horatio Bond, chief engineer of the National Fire Protection Association, visited London for four months in late 1942 to give advice on large-scale fire destruction, followed a few months later by James McElroy, another senior NFPA engineer, who worked for the rest of the war at the RE8 headquarters at Princes Risborough in Buckinghamshire, first producing so-called fire division maps of German cities, showing each urban “cell” that needed to be set alight, then going on to produce vulnerability maps of major industrial targets for the Eighth Air Force Operational Research Section, to show what proportion and what type of incendiary bomb should be carried by aircraft.¹¹⁵

The American air force was impressed by British incendiary practice. In April 1943, Arnold spelled out for his assistant chief of staff for matériel the ways in which the Eighth Air Force would be expected to use incendiary bombs: first, for burning down precise industrial targets when a greater degree of destruction was possible using fire; second, by starting fires “in the densely built-up portions of cities and towns” to create a beacon for RAF attacks at night on the same city or town; third, for burning down densely built-up areas “when the occasion warrants.”¹¹⁶ American forces used principally the 4-pound incendiary (renamed the M-50), the M-17 110-bomb cluster, the M-47 70-pound oil and rubber bomb, and the M-76 473-pound oil bomb. Extensive plans were laid in 1942 for incendiary bomb production, 39 million bomb cases in that year, 107 million in 1943, divided between the different services and Lend-Lease supplies for Britain.¹¹⁷ Extensive research on the effects of incendiary bombing resulted in a set of recommendations for the Eighth Air Force in September 1942, which differed from RAF practice principally by preferring the 473-pound oil bomb over the 4-pound incendiary and identifying weather as a vital condition: “A high wind is still the best weapon.” Bomb patterns were to be dense enough “to guarantee real conflagrations.”¹¹⁸ Extensive experiments were carried out in Britain on the use of oil bombs of different weights, and in April large quantities were finally shipped to American bases for operational use. They were first employed extensively in July 1943, the month of Operation Gomorrah. By the end of the war the Eighth Air Force had dropped a total of 90,357 tons of incendiaries and incendiary clusters, totaling 27 million 4-pound bombs and 795,000 heavy incendiaries.¹¹⁹

The two-thirds load of incendiary bombs aimed at the residential areas in Hamburg north of the river Elbe to try to stimulate an uncontrollable conflagration was thus no accident. The choice of Hamburg is not difficult to explain. The Air Ministry Target Committee in April 1943 defined Hamburg as “No. 1 priority” because of its shipbuilding industry.¹²⁰ It had been attacked repeatedly since 1940, usually in small raids that inflicted light damage on Germany’s second-largest city. There were 127 small raids between 1940 and the end of 1942; the 10 raids in 1943 before Gomorrah resulted in just 142 deaths and the destruction of 220 buildings.¹²¹ Hamburg was fourth on the MEW list, behind Berlin, Duisburg, and Bochum, all three of which had been attacked in the late spring. The MEW rankings in the “Bomber’s Baedeker” had twenty-one industrial targets ranked 1+ or 1 and twenty-five ranked 2, substantially more than most other cities.¹²² The vulnerability of Hamburg was magnified by its proximity to British bases, the conspicuous coastline, and its sheer size. In a survey of German cities already subject to incendiary attack, the ratio of incendiary to high-explosive damage in Hamburg was judged to have been 13:1, higher than any other except the nearer port of Wilhelmshaven.¹²³ It is not clear whether Harris was ever shown these figures, but his desire to launch a spectacular operation fitted with the geographical pattern of the new attacks and Air Ministry ambitions. Not to have attacked Hamburg in force would have been more difficult to explain.

The decision was nevertheless not easily endorsed and rested in the end on Churchill’s approval. The chief argument concerned the decision of whether to use a simple tactical device to temporarily blind all German radar, known by the code name Window. The tactic consisted of dropping very large quantities of small aluminized strips that would create a confused blur of echoes on enemy radar screens and make accurate detection of the bomber stream impossible. It was first developed in late 1941 by Robert Cockburn at the Telecommunications Research Establishment in Malvern, and early trials led to the manufacture of the material with a view to using it against German radar in May 1942. Frederick Lindemann persuaded Portal to cancel the instruction, and a fresh set of tests showed that the latest British ground and airborne radar was highly susceptible to Window if the Germans retaliated with it. Lindemann insisted that it should not be used operationally until an antidote could be found. The Germans did develop the same technique, known as Düppel, but, like the RAF, hesitated to develop and use it. However, by late 1942 new British AI radar for night fighters and GCI (ground-controlled interception) radar for ground control, together with a new American airborne radar, SCR-270, all had the capacity to survive a Window assault. American experiments also showed that maximum effect would come from a large number of narrow strips, about half as long as the wavelength of the German Würzburg radar, which controlled the German air defenses. Window was manufactured in 30-centimeter (12-inch) strips, 1.5 centimeters (0.6 inch) wide, with paper on one side and aluminum foil on the other, in bundles of 2,000 for release from each aircraft in the bomber stream.¹²⁴

The combination of improved British defenses and the weakness of the German bomber arm in the west finally persuaded Portal to approve the employment of Window, but now the chiefs of staff objected to its operational use before the invasion of Sicily in early July, in case the German Air Force used it to confuse Allied air support during the landing. As a result its use was postponed until Operation Gomorrah, following approval by Churchill on July 15.¹²⁵ A few days later Churchill was faced with another objection, this time the choice of Hamburg as a target at all. Henry Tizard, the government’s chief scientific adviser, wrote to Churchill and Portal deploring the planned destruction of Hamburg on the grounds that it would be a useful capital for the Allies to occupy when they ran postwar Germany, and that its population was “anti-Russian, anti-Prussian and anti-Nazi,” and might soon be “anti-war.” Churchill sent the letter to the chiefs of staff, but Portal had already answered Tizard, explaining that Hamburg was too important a target to ignore. “It is a moot point,” he continued, “whether bombing produces a more desirable effect when directed upon anti-Nazis than upon the faithful,” but he was content for Harris to find out whether Hamburg’s anti-Nazi sentiment would be stung into action by bombing. Churchill took this for approval.¹²⁶ By then the first of the Gomorrah raids had already taken place.

The operation against Hamburg, like the heavy bombing of Cologne in late June and early July, was spread over ten days from the first RAF raid on July 24–25 to the final raid on August 2–3.¹²⁷ The opening night-bombing raid by 728 aircraft was the first to use Window. Around eighty miles from the target the Pathfinder Force and the main force that followed emptied bundles of foil strips at the rate of one bundle a minute. The strips worked perfectly, creating numerous echoes on the cathode-ray radar screens and presenting German night fighters with a confusion of false information. Searchlights and antiaircraft artillery had to improvise a barrage of light and fire in the hope that bombers might be deterred anyway. In just under one hour 2,284 tons of bombs were dropped, including an average of 17,000 incendiaries for every square kilometer.¹²⁸ Although fewer than 50 percent of the bombers hit the three-mile aiming zone, the rest hit the large central and northwestern residential districts, killing, according to the Reich Statistical Office, 10,289 people, three times more than the worst raid so far.¹²⁹ Over the following two days the Eighth Air Force attacked targets by day in northwest Germany. On July 24, 218 bombers bombed shipbuilding targets in Hamburg and Kiel, losing 19 aircraft in the process; on July 25, 96 aircraft attacked Hannover and 54 raided Hamburg, with the loss of a further 18 aircraft, a rate over the two days of more than 10 percent, an indication that the warnings about the dangers of daytime bombing voiced at Casablanca had not been misplaced. The two American raids killed 468 people in Hamburg.¹³⁰

The RAF raid that followed on the night of July 27–28 was a textbook example of the incendiary planning of the previous two years. It was helped by the prevailing meteorological conditions. There had been rainfall on July 22, but the remainder of the week was dry. Humidity levels, which had been high earlier in the month, fell abnormally, reaching 46 percent on July 25 and only 30 percent on July 27. Temperatures soared in the last week, reaching 32°C (90°F) on the evening of July 27; low humidity and high temperatures remained over the following two days. These summertime conditions favored the chances of a major conflagration.¹³¹ The Pathfinder Force dropped markers several miles east of the center of Hamburg, but the 729 aircraft concentrated their 2,326 tons well on the packed working-class districts of Hammerbrook, Borgfelde, Hamm, Billwerder, Hohenfelde, and Rothenburgsort. The raid lasted just over an hour. The concentration of approximately 1,200 tons of incendiaries on an area of two square miles created numerous major fires that soon merged together into a roaring inferno. Water shortages caused by the earlier heavy raid hampered firefighting. Many emergency workers and vehicles were farther west in the city still coping with the aftermath of the first fires where the civil defense control room had been destroyed. Efforts to stem the fires proved useless. What followed, in the words of Hamburg’s police president, was a “hurricane of fire . . . against which all human resistance seemed vain.”¹³² The illusion of a hurricane was caused by the scale and intense heat of the conflagration, which caused fire winds that drove the flames across natural firebreaks. The inferno created a pillar of hot air and debris that rose quickly to a height of more than two miles above the city. Greedy for more oxygen, the fire drew in cold air from the surrounding area with such force that the new winds reached hurricane-force strength in the area of the fire, collapsing buildings, uprooting trees, and sucking human bodies into the flames where they were swiftly incinerated or mummified. Acting like giant bellows, the winds created temperatures in excess of 800°C that destroyed everything combustible barring brick and stone. Oxygen was sucked out of the thousands of basement and cellar shelters, leaving their inhabitants to die slowly of carbon monoxide poisoning.¹³³ An estimated 18,474 people died during the night. An area of more than twelve square miles was burnt out.

This was not the end for Hamburg. Harris’s intention of destroying the city brought two more major raids on July 29–30 and August 2–3. The first, carried out by 707 bombers, dropped a higher tonnage than the firestorm night, but did not create a second thermal hurricane. Large residential areas were again burnt out and an estimated 9,666 killed. The final raid in August was abortive. Large thunderstorms protected Hamburg. The Pathfinder force failed to mark the city and most aircraft dropped their bombs over northern Germany or in the North Sea. The final raid killed 78 people. The cumulative total for Operation Gomorrah calculated by the local police authorities by November 1943 was 31,647; the figure was revised in May 1944 to 38,975, which is close to the figure currently favored of 37,000.¹³⁴ Around 900,000 people evacuated the city and 61 percent of Hamburg’s houses and apartments were destroyed or damaged, together with 580 industrial premises and 2,632 shops. The RAF bomber force lost only 87 aircraft, or 2.5 percent of all sorties, thanks partly to the use of Window, partly to the shock effect on Hamburg’s defenses. The post-raid report noted that smoke obscured much of the evidence of destruction but confirmed that the “amount of residential damage is very great.”¹³⁵

The consequences for the German air defense system were profound. The catastrophe at Hamburg even more than the raids on the Ruhr catapulted the German Air Force into abandoning the principle of the fixed front, embodied in the Kammhuber Line and heavy reliance on antiaircraft fire, and substituting instead a much larger fighter force based on new forms of combat. Even before the conflagration at Hamburg, Göring, despite his anxiety to begin offensive air operations against Britain, admitted to his staff that “air defense is now in my opinion decisive.” The terrible events in Hamburg accelerated the shift to priority for fighter production and intense efforts to combat the scientific lead in radar and radio countermeasures that the Allied air force had established over the summer months.¹³⁶ By October, Göring’s deputy, Erhard Milch, had drawn up firm plans for the output of more than 3,000 fighter aircraft a month, and more speculative plans for 5,000 fighters every month during 1945, at the expense of further bomber production.¹³⁷ By simplifying production methods, reducing the large number of aircraft types, and abandoning the habit of regular modification, Milch calculated that the output could be achieved without a large increase in labor.¹³⁸ He had already taken over central responsibility for radar and radio development in May and had begun an immediate acceleration of research and production. In July a new office for high-frequency research was set up under Dr. Hans Plendl, which drew in 3,000 scientific personnel to combat the variety of devices that had fallen into German hands during the Ruhr battle, thanks to the recovery of equipment from crashed bombers.¹³⁹

After the success of Window, priority was given by the Germans to developing radar that would be more immune to its effects in order to increase as rapidly as possible the hitting power of the German fighter force. Permission was now given to build up units trained for Hermann’s Wilde Sau technique, while on the day following the final heavy raid on Hamburg, July 30, the Zahme Sau tactic of controlled infiltration of the bomber stream was approved by Milch and Hubert Weise, overall commander of the German home-front air defense. The result was a substantial diversion of guns and aircraft away from the fighting fronts where they were needed more than ever by the summer of 1943. By late August there were over 1,000 fighter aircraft stationed in Germany, 45.5 percent of all German fighter strength, and a further 224 in northern France. Over the same period the number of heavy antiaircraft guns on the home front increased from 4,800 before Gomorrah to over 6,000 by the end of August, including more of the heavier 105-millimeter and 128-millimeter models. Shortages of skilled personnel hampered the antiaircraft effort throughout the year, and for the first time substantial numbers of women and young people were recruited as Flakhelfer; on Hitler’s orders training began in August 1943 for 250 units of Reich Labor Service boys to undertake air defense duty in the year before they joined the armed forces.¹⁴⁰

In Britain the Hamburg raids were treated as a great success. The director of intelligence in the Air Ministry thought that the operation confirmed the superiority of incendiary over high-explosive bombs: “The complete wipe-out of a residential area by fire is quite another and better conception. May it long continue!”¹⁴¹ Robert Lovett wrote to Eaker from Washington that the War Department was keen to see the photoreconnaissance images of Hamburg’s destruction as soon as possible: “The pasting Hamburg got must have been terrific.” Eaker replied that the raids had “a tremendous effect.”¹⁴² For Harris, Operation Gomorrah was more than he could have hoped for. The post-raid assessments, now based on RE8 research on man-months lost and acreage destroyed, transformed the statistical image of Bomber Command’s achievements and gave Harris the weapons he needed to argue the case that bombing was capable of knocking Germany out of the war. If there had been a defensiveness about Bomber Command strategy before Hamburg, the new evidence could be used to strengthen Harris’s hand. On August 12 he wrote to Portal, who was with the Combined Chiefs at the Quadrant conference in Quebec, that in his view the bombing war was “on the verge of a final show-down.” Harris was certain that with the same concentration of effort “we can push Germany over by bombing this year.”¹⁴³

The results of Operation Gomorrah could only be assessed by the crudest measurements, since the detailed effects were not yet known even to the German authorities. This meant chiefly an assessment of the acreage of Zones 1 and 2A, the inner-city residential areas, “devastated” predominantly by incendiary attack. The Hamburg raid had the effect of doubling the amount of damage inflicted on German cities so far. Up to the end of June it was estimated that 12.27 percent of the inner-city areas had been destroyed; by the end of September this figure had increased to 23.31 percent, 18,738 acres against 9,583.¹⁴⁴ The intelligence staff at Bomber Command headquarters used this material to present a statistical way of measuring success based on the following three criteria:

Tons of bombs claimed dropped per built-up acre attacked = “Effort”

Acres of devastation per ton of bombs claimed dropped = “Efficiency”

Acres of devastation per acre of built-up area attacked = “Success”

On this basis it could be demonstrated that “success” (acres destroyed per acres attacked) had increased from 0.001 at the end of 1941, 0.032 at the end of 1942, to 0.249 at the end of October 1943. The acres of Germany’s central urban area devastated had increased by a factor of 24 in the course of 1943.¹⁴⁵ These figures gave no indication of what effect this was likely to have either on Germany’s war effort or on the state of mind of those bombed. The MEW warned the Air Ministry that it was hard to judge German conditions. They estimated that Operation Gomorrah had cost Hamburg the equivalent of 1.25 million man-months, or 12 percent of the city’s annual production. The cumulative effect of all Bomber Command’s attacks during the late spring and summer was estimated at a more modest 3 percent of Germany’s potential production effort.¹⁴⁶

Evidence soon became available to show that the damage was not as crippling as had at first been hoped. The port of Hamburg, which had been less severely damaged than the city, was estimated by MEW to be operating at 70 percent of its capacity again by the end of August; intelligence on the Blohm & Voss shipyards, one of the principal targets for the Eighth Air Force, suggested that they had not been destroyed and were still functioning.¹⁴⁷ By November the city was back to 80 percent of its pre-raid output. When Göring was captured at the end of the war with a train full of possessions, the American army found among them a presentation folio of more than a hundred charts and graphs on the remarkable recovery of Hamburg.¹⁴⁸ The one statistic that could only be guessed at was the number of dead. This was not a measurement used by Bomber Command, whose calculations of “effort,” “efficiency,” and “success” became increasingly more abstract as the war went on. The acreage destroyed was released to the British press as early as August 6, though the RAF communiqué emphasized that the damage “particularly covers the principal manufacturing districts and the docks and wharves.”¹⁴⁹ Sources in Sweden suggested the figure of 58,000 dead, which was reprinted in British newspapers later in August. News of the firestorm became available from a correspondent of the Swiss paper Basler Nachrichten only in November 1943, when it was circulated among the antibombing lobby in Britain and published in the New Statesman. The claims that 20,000 bodies had been found, incinerated to a degree not even found in a crematorium, were contested by British scientists who were shown the material. They dismissed the description of the firestorm as scientific nonsense, an “intrinsic absurdity.”¹⁵⁰

Stalemate over Germany

In the aftermath of Operation Gomorrah both Allied bomber forces expected a growing success rate and both were impatient to achieve it. There was a growing confidence that the aims laid down at Casablanca might now be operationally in their grasp. Lovett wrote to Eaker in early July about the wave of optimism surging through the United States “that the Air Forces will knock Germany over by Christmas.”¹⁵¹A report on the combined offensive produced by Eaker in early August 1943 talked of knocking out the industrial props of Germany one by one until “the German military machine comes closer and closer to collapse.”¹⁵² Harris telegraphed Portal at the Combined Chiefs of Staff conference in Quebec in mid-August that the combined efforts of the two forces should be enough, once again, to “knock Germany stiff.”¹⁵³ Portal was keen for Harris to attack Berlin with the same force as Hamburg: “In present war situation,” he telegraphed from Quebec, “attacks on Berlin on anything like Hamburg scale must have enormous effect on Germany as a whole.” Harris explained that this would need 40,000 tons of bombs and good weather, but Berlin was to be next on his list.¹⁵⁴

For both bomber forces there was a race against time. The Eighth Air Force needed to be able to show that it was meeting the Pointblank requirement to undermine the German Air Force in time for the projected invasion of France in May 1944. Whatever the American public expected of the bombing campaign, Eaker’s directive was to use airpower to prepare the way for the ground armies, and the progress of the combined offensive was regularly measured against this requirement. The Quadrant conference in Quebec in August 1943 reiterated that destruction of German airpower was to have “the highest strategic priority.”¹⁵⁵ By October it was evident that the German fighter force was growing in strength and that the efforts to reduce it had been ineffective. The Eighth Air Force was directed to speed up its assault on its list of essential German targets; out of 128 attacks on Europe, only 50 had so far been against Germany.¹⁵⁶ Throughout the last weeks of 1943 and the first months of 1944, Eaker, and his successor in January, Brigadier General James Doolittle, were told insistently by Arnold and Spaatz that Pointblank “must be pressed to the limit.” In January 1944 the chiefs of staff wanted the offensive to focus only on German fighter strength during the preparatory period for Operation Overlord, as the German Air Force had been asked to do before Sea Lion in 1940.¹⁵⁷

Harris and Bomber Command ran a different race. He wanted city bombing to bring the war to a conclusion without an extensive and costly ground invasion, and this meant doing such severe damage to Germany’s urban population and environment in the months after Gomorrah that the German war effort would crumble. A joint report on the progress of the combined offensive drawn up in November 1943, but clearly influenced by Harris, played down the impact of American raids and highlighted the significance of area attacks on industrial cities, which had already reduced German war potential, it was claimed, by 10 percent and “may well be fatal” if the figure could be doubled in the next few months.¹⁵⁸ Harris drew up a list the same month of the different urban target areas (with Berlin the priority) for “the continuation and intensification” of the offensive, each city defined as “largely destroyed,” “seriously damaged,” “damaged,” or “undamaged,” and with the added hope that the Eighth Air Force would soon join in the bombing of the German capital.¹⁵⁹ In December, Harris used the urban damage figures in a report to Portal and Sinclair in which he claimed that the physical destruction of 40–50 percent of the urban area of the principal towns of Germany would produce, by April 1944, the month before Overlord, “a state of devastation in which surrender is inevitable.”¹⁶⁰

The two ambitions were not easily compatible, and in the event difficult to achieve. Over the course of the late summer and autumn both forces carried out major raids that became large-scale air battles of increasing severity and intensity as German defenses were strengthened and the German Air Force was freed at last to pursue more effective operational tactics. On August 12, 133 B-17 Flying Fortresses attacked Bochum in the Ruhr-Rhineland and lost 23 out of 133 aircraft (17 percent of the force). The next major battle came on August 17, deliberately chosen as the anniversary of the Eighth Air Force’s first mission in 1942. The object was to inflict a spectacular blow. The targets selected were the ball-bearing works at Schweinfurt and Messerschmitt Me109 fighter production at Regensburg. For this raid American bombers were divided up; those destined for Regensburg under the command of Curtis LeMay would continue on to North African bases, those for Schweinfurt, commanded by Brigadier General Robert Williams, had to fight their way there and back. Both cities were the farthest the Eighth Air Force had yet flown into German airspace and involved the largest number of bombers so far dispatched, a total of 376 B-17s—a reflection of a sudden escalation in the supply of both crews and aircraft. The Messerschmitt works produced 18 percent of all Messerschmitt Me109 production; the ball-bearing works at Schweinfurt produced 45 percent of the supplies of German ball bearings. The aircraft set off in the late morning of August 17 and were attacked by German fighters from the moment Allied fighter cover ended, at Antwerp on the Channel coast. The group destined for Regensburg reached the target near midday and dropped 298 tons, killing 400 people and causing the temporary loss of 20 percent of fighter output. On the way there 12 B-17s were shot down, and 12 more were lost in the aftermath as they flew on across the Alps to Tunisia. Poor repair facilities in North Africa left more bombers grounded, and after a long return flight with further losses, only 55 out of the original 146 returned to English bases.¹⁶¹

The First Bomber Division faced an even greater battle. For more than three hours in the afternoon flight the bomber stream was subjected to persistent fighter attack, losing a total of 36 aircraft. Around three o’clock, 424 tons of bombs were dropped on the Schweinfurt works, killing 141 people, destroying two works completely, and seriously damaging a number of others. The German Armaments Ministry recorded a temporary loss of 34 percent of ball-bearing production, though large reserve stocks were available to cushion the blow.¹⁶² For the Eighth Air Force the cost for just two targets on their list was exceptional. Together with the 60 aircraft shot down, 176 were damaged and 30 remained in North Africa. Including those lightly damaged, the casualty rate was 71 percent; counting those destroyed, severely damaged, or out of theater, the loss rate was 31 percent, levels that could scarcely be sustained for more than a few more raids. The German Air Force lost 28 fighters to the concentrated fire of the B-17s, though in common with almost all air-to-air engagements, American aircrew claimed to have shot down a remarkable 288 enemy aircraft.¹⁶³ In September a raid on Stuttgart, undertaken while Arnold was visiting Britain, saw the loss of 65 bombers (19 percent) out of a force of 338, with little damage to the city itself.

The Eighth Bomber Command nursed its wounds for more than a month before beginning a new series of deep-penetration attacks in October. A raid on the coastal city of Anklam on October 9 cost 18 bombers out of a force of 106, but the most famous battle of the bomber war took place during a second raid on Schweinfurt on October 14, when 65 bombers were lost out of an attacking force of 229, a total loss rate of over 28 percent of the force. Fighters accompanied the aircraft as far as Aachen, thanks to the addition of extra fuel tanks, which pushed their range to 350 miles, but after that the bomber stream was subjected to hundreds of attacks with rocket and cannon fire from enemy fighters. Eaker wrote to Arnold after the raid, “This does not represent disaster; it does indicate that the air battle has reached its climax.”¹⁶⁴ But so severe was the risk on any raid past fighter cover that for the next four months operations were only carried out with increasing numbers of bombers on cities within easy range. As a result, Kiel, Bremen, Wilhelmshaven, and Emden—already the victims of the first year of Bomber Command raids thanks to their proximity—became the recipients of occasional heavy bombing, while the majority of American attacks now took place once again over France. The Eighth Air Force, like the German Air Force and Bomber Command years before, began to think about the possibility of night attacks, and in September 90 B-17s were converted for practice and training in night flying.¹⁶⁵ In three months of raiding, the Eighth Bomber Command lost 358 B-17s in combat, suffering the highest loss rate of the war in October. The campaign did not come to a complete halt, but the next raid deep into Germany was made only on February 20, 1944, in very different circumstances.¹⁶⁶

Bomber Command also began to experience higher losses and, despite the advent of the target-finding apparatus Oboe and H2S, continued to hit urban targets with intermittent success. The rate of expansion in the first half of the year was not sustained: operationally ready aircraft within the command increased by more than two-thirds between February and June, but only by another quarter between July and December. Pilot strength was 2,415 in June, 2,403 in December.¹⁶⁷ On August 17–18, Harris was ordered to mount a more precise attack on the German Air Force research station on the Baltic coast at Peenemünde. The raid by 560 aircraft dropped 1,800 tons mainly of high explosive on the research stations and accommodations (also killing 500 Polish forced laborers in a nearby labor camp). Extensive damage was done to the rocket research program in one of the few “precision” raids attempted by the command in 1943, but 40 aircraft were shot down, a loss rate of 6.7 percent. Harris then shifted his focus to Berlin. The first raid on August 23–24 missed the center of the capital and hit the southern suburbs, killing 854 people and destroying or damaging 2,600 buildings. The loss rate of 7.9 percent was the highest on any raid since the start of the offensive. A second raid on August 31–September 1 cost 7.6 percent of the attacking force, while H2S marking was erratic, causing the bomber stream to drop bombs up to thirty miles from the aiming point. Only 85 houses were destroyed in Berlin and 68 people killed. So heavy were the losses to Stirling and Halifax squadrons that the third attack, on September 3–4, was made only with Lancasters, but not only did the H2S marking once again miss the main aiming point, but 7 percent of the Lancaster force was lost. The first “Battle of Berlin” petered out until November in favor of less dangerous targets.¹⁶⁸ On the night of October 22–23 a second firestorm was created in the small city of Kassel, where H2S marking was for once accurate. The raid report noted that the whole city area “was virtually devastated.”¹⁶⁹ The death of an estimated 6,000 people was a higher percentage of the city population than in Hamburg. Some 59 percent of Kassel was burnt out and 6,636 residential buildings destroyed. Armaments production was hit heavily by the destruction of the workers’ quarters, but like Hamburg’s it revived after two or three months to around 90 percent of the pre-raid level.¹⁷⁰ The raid put Kassel for the moment at the top of Bomber Command’s list of the proportion of city buildings destroyed or damaged, ahead of Hamburg, 58 percent to 51 percent.¹⁷¹ But Bomber Command lost 43 aircraft, 7.6 percent of the force, a steady attrition of its strength. Over the course of 1943 the command lost 4,026 aircraft, 2,823 in combat.¹⁷²

The high losses of the autumn months created a growing sense of urgency and uncertainty in both bomber forces, despite the brave face turned to the outside world. The claims that had been made from Casablanca onward served the cause of service politics as well as Allied strategy; the stalemate developing in late 1943 as the balance shifted away from offensive toward defensive airpower exposed the bomber force to the risk of relative failure. Although both forces advertised their success in diverting ever-increasing numbers of German fighters to the defense of the Reich, this was in some sense a Pyrrhic victory, since the bomber forces were now subject to escalating and possibly insupportable levels of loss and damage. Lovett reminded Eaker in September that if Germany did not collapse over the winter, the public would think “our ‘full-out’ offensive doesn’t work.”¹⁷³ A paper prepared for Portal in October argued that a failure to demonstrate what strategic bombing could do would have “dangerous repercussions upon post-war policy.”¹⁷⁴ Both forces wanted the other to help more in pursuit of what was seldom a common ambition. In October, Portal told Eaker and Harris that the Pointblank aim to eliminate the German Air Force as a preliminary to a more sustained offensive had so far failed. “Unless the present build-up of the G.A.F. fighter force is checked,” he continued, “there is a real danger that the average overall efficiency and effectiveness of our bombing attacks will fall to a level at which the enemy can sustain them.” The operational evidence showed that Bomber Command had so far devoted only 2 percent of its effort to fighter aircraft assembly plants (as by-products of area attacks), and Harris was now instructed to launch operations against six principal cities associated with fighter production.¹⁷⁵

Harris treated the request as he had treated other “panacea” targets. In July he had been briefed to attack Schweinfurt as the most vulnerable link in the chain of German war production. Alongside claims for its importance he scribbled “sez you!” in the margin, and no night raid was made. In December he was told again that his command should bomb Schweinfurt at night. He rejected the idea on the grounds that his force even now could not hit a small city with any certainty, and would not waste time on a single target when there were bigger cities to destroy and “only four months left!”¹⁷⁶ To the Air Ministry’s complaints that Bomber Command should have done more to support the Eighth Air Force campaign against German airpower, Harris retorted that he would not do their job for them. The two forces could not be regarded as complementary, Harris continued, since his command had dropped 134,000 tons in 1943 and the Eighth Air Force only 16,000, much of it on less important targets.¹⁷⁷ When RE8 researchers suggested that Bomber Command attacks had by their estimate actually done little more than reduce German economic potential by 9 percent in 1943 (even this figure turned out after the war to have been overoptimistic), Harris responded angrily that in the cities his force had devastated the proportion must self-evidently be higher. Even this devastation, as the Air Ministry reminded him, had affected only 11 percent of the whole German population, but Harris was a figure difficult to gainsay even over an issue of real strategic significance.¹⁷⁸ He stuck to area bombing in preference to selected targets and as a result made Pointblank harder to achieve, while paradoxically contributing to the stalemate he was trying to break.

While the two forces argued over priorities, the greatest battles of the bomber war were being played out in the skies over Germany. In all cases the majority of bombers succeeded in carrying out their bombing mission; there was never a point when the bomber did not get through. Although the bomb loads were often spread over areas widely distant from the actual target, wherever they landed they did serious damage to the home population, in town or countryside, on which they fell. These aerial battles were nevertheless distinct from the bombing itself. They were fought between the enemy defensive forces—guns, night- and day-fighter aircraft, searchlights, decoys, and barrage balloons—and the intruding enemy. At this stage of the war the problems facing both sides, though they were supported by increasingly sophisticated scientific and technical equipment and weaponry, remained the same as they had been since the onset of the bombing war in 1940: weather conditions, bombing accuracy, the balance between defensive and offensive tactics, operational organization, and force morale. These factors profoundly affected what the two bomber commands were capable of achieving.

In an age where radar aids to navigation and electronic guidance systems were in their infancy, weather continued to play an arbitrary and intrusive role in the conduct of air warfare. The final United States Bombing Survey report on weather effects regarded them as “a major controlling factor” in the operation of Allied bomber forces. For the Eighth Air Force, which relied more on visual bombing, weather prevented any operations for a quarter of the time, while 10 percent of all aircraft that did take off aborted because of weather conditions. It was calculated from unit records that the operational rate of the Eighth Air Force was only 55 percent of the potential effort because of the effects of northern Europe’s rainy climate. Low cloud and fog were the main culprits.¹⁷⁹ Calculations by the Army Air Forces’ director of weather in late 1942 found that the average number of days per month when the sky over the target was absolutely clear was 1 or 2 in the winter months, rising to a peak of 7 in June, a total of 31 days in the year. Days when the major limiting factors of high wind, ice, or more than three-tenths cloud cover were absent were more numerous, but still numbered only 113 out of the year, again with a low of 6 in the winter months and a peak of 12 in June.¹⁸⁰ The air force weather service developed a sophisticated pattern of weather prediction, based on the experience of the burgeoning civilian airline business in the 1930s, providing regular climate data, forecasts of current weather trends, and a precise operational forecast for particular missions.¹⁸¹ Even with reasonably accurate forecasting, weather conditions could alter rapidly and unexpectedly. The combat diary of the 305th Bombardment Group in 1942–43 can be taken as an example: November 23, Lorient “covered by cloud”; December 12, Lille, “cloud cover at the target”; January 23, Brest “obscured by the cloud cover”; February 4, Emden, “no bombs were dropped due to the clouds”; and so on.¹⁸²

Bomber Command was less affected by the weather because area bombing could be carried out in weather conditions that were less than ideal, but British experience also showed that “average good visibility” was only available between five and nine nights in the summer months and three to five during the winter.¹⁸³ To the end of the war, Harris continued to cite the weather as a principal explanation for why Bomber Command could not switch to attacks on precise target systems.¹⁸⁴ With the arrival of electronic aids to navigation, bombing could be carried out through cloud and smoke, though the return to a base suddenly shrouded in fog caused regular accidents. Weather did make severe demands on pilots as they struggled at night to cope with the elements, as the following recollection of the last night of Operation Gomorrah illustrates:

We set course north to the targets and flew into thunderstorms resulting in heavy icing. We could hear the ice breaking off the sides of the aeroplane and the propellers, and then we were losing power virtually on four engines with heavy icing and I then lost control of the aeroplane. . . . It started to descend with the weight of ice and at that stage we were hit by anti-aircraft fire. . . . The aeroplane by this stage was completely out of control with icing . . . a very frightening experience.

On this occasion pilot and crew survived.¹⁸⁵

Of all the problems posed by the generally poor weather conditions over Germany, maintaining adequate levels of bombing accuracy was the most significant. For the Eighth Air Force, clear weather was essential if precise targeting was to survive. In ideal conditions the early raids on France showed an average error range of 1,000 yards, but conditions were seldom ideal. An investigation carried out in February 1943 highlighted problems of poor accuracy and, alongside weather and enemy defenses, blamed poor onboard teamwork and inexperienced pilots. It was found that the method of bombing in squadron (or group) formation reduced accuracy substantially for the last formations to bomb because of the smoke and fire on the ground; up to July 1943 an average of only 13.6 percent of bombs fell within 1,000 feet of the aiming point, but for the last formation to bomb the figure fell to only 5 percent.¹⁸⁶ Anderson pressed his commanders to focus all their efforts on putting more bombs “on the critical points of the targets.” “It is evident,” he continued, “that our bombing is still not up to the standards that it can be.”¹⁸⁷ Cloud, industrial haze, and smoke screens made this inevitable. The American offensive began without a Pathfinder Force and no electronic aids to navigation. In March 1943, Eaker had asked Portal to supply both the Oboe and H2S equipment for use by the American formation leaders. Portal offered equipment for only eight aircraft and training facilities for the crews. By the early autumn there were three Pathfinder units forming, one with H2S, two with the American version of the technology, known as H2X.¹⁸⁸ Eaker found the temptation irresistible to use the new guidance system to allow bombing when the target was obscured. In September 1943 the Eighth Air Force undertook its first deliberate blind-bombing attack on the German port of Emden. The bombing was scattered around Emden and the surrounding region, but the fact that it was now possible to arrive over the area of the target in poor weather introduced the American air forces to area bombing.

Escalation in this case was dictated by the technical impossibility of bombing accurately for more than a few days a month. From September 1943 onward, American bombers were directed to attack city areas through cloud in the hope that this would hit the precise targets obscured by the elements. Sensitive to opinion, the raids on city areas were defined, as with Bomber Command, as attacks on industrial centers or, increasingly, as “marshaling yards.”¹⁸⁹ The distinction for most, though not all, American raiding from British area bombing was intention. The civilian and the civilian milieu were never defined as targets in their own right, even if the eventual outcome might make the distinction seem merely academic. Around three-quarters of the effort against German targets between 1943 and 1945 was carried out by blind bombing using H2X equipment; from October to December 1943 only 20 percent of bombs dropped using radar aids came within five miles of the aiming point, an outcome little different from Bomber Command’s in 1941–42.¹⁹⁰ High levels of accuracy were reserved for primary industrial targets in good weather, where accuracy levels increased from 36 percent within 2,000 feet in July 1943 to 62 percent in December. American practice became one of selective precision when visibility was good, and less discriminate attacks when visual conditions were poor.¹⁹¹ The result, as with Bomber Command, was to increase the number of German civilians killed and houses destroyed by American bombers, which now carried much higher incendiary loads for blind-bombing raids. Even precision bombing resulted in widespread damage to the surrounding civilian area. The USAAF in-house history of radar bombing was more candid than the public image: “Neither visual nor radar bombing ever achieved pin-point bombing; both methods were, in effect, methods of area bombing in the sense that a certain percentage of bombs fell within an area of a certain size, the rest falling without.”¹⁹²

The problem for Bomber Command was different. Here unacceptably inaccurate bombing had to be improved to achieve higher concentrations on the chosen urban areas. The introduction of Oboe and H2S contributed to raising the average accuracy of the attacking force, but concentration could be lost if the weather deteriorated or the Pathfinder Force missed the aiming point, or German decoys were successful in diverting a proportion of the attacking force. In raids using Oboe in 1943, the number of aircraft that bombed within three miles of the target ranged from 77 percent against Cologne in July to 32 percent against Bochum in September, but poor weather prevented evaluation of at least half the raids. Oboe certainly proved to be the more successful of the two methods but could only reach as far as the Ruhr-Rhineland. For H2S operations the scale of accuracy was even longer: 86 percent against Kassel in October to 2.1 percent against Berlin on August 31. The average for the twenty-three H2S raids that could be plotted was 32 percent, a substantial improvement on the Butt Report evidence, but still a low level of concentration.¹⁹³ Research on the first large-scale blind-bombing raid using H2S in poor weather, against Mannheim-Ludwigshafen on November 17–18, estimated that perhaps 60 percent of the attacking aircraft had hit the conurbation itself. These figures indicated that the force had at last adopted a technology and tactics that might reduce the amount of wasted effort. This was partly due to the additional training for Pathfinder navigators organized by the Bombing Development Unit, set up in late 1942, consisting of flights over British cities. These simulated raids showed wide deviation from the putative aiming point, but an average of 50 percent of “hits” within a three-mile radius (four miles for London). The trials showed that H2S worked well over certain urban targets, but poorly over sprawling urban areas or cities surrounded by hills, as had already been found over Germany. The Operational Research Section of Bomber Command calculated that this was probably the best to be hoped for when bombing cities. Improved though Bomber Command accuracy was from the poor state of 1941, every second bomb was still miles from the aiming point.¹⁹⁴

The problems of weather and bombing accuracy highlight a factor about the bombing war that is seldom given the weight it deserves in assessing what was and was not possible for the forces at the Allies’ disposal. The operations mounted week after week during the last two years of war were of unprecedented scale and complexity, employing some of the war’s most sophisticated equipment. For both air forces the pre-raid preparation required all the conventional demands of battle—tactical, logistical, and technical; the intelligence and operational research reports had to be factored into each calculation and the weather closely monitored. Organizing a raid with hundreds of aircraft coordinated over long distances promised all kinds of hazards. Pilots had to create formations while avoiding accidents and to synchronize their flights with the escorting fighters as far as their range would allow. Along the target run and over the target itself there were precise instructions about combat, evasion, target recognition, heights, and speeds. On return there was the process of debriefing crews, coping with casualties, and estimating the outcome of the raid. Each battle was self-contained, but for commanders and their crews the campaign was continuous, more so than for almost any other form of combat over the four or five years of war.

The typical instructions for a Bomber Command raid illustrate the close attention to planning detail and the range of demands made of the crews. Aircraft from four or five bomber groups were given instructions about force size, composition of the bomb load, routes to the destination target (or for the decoy attacks), and the timing of each of five waves of attacking aircraft, which had to drop their bombs within a twenty-minute period to maximize impact. There were instructions on Window, on the Mandrel jammer, radar and wireless use, and the chosen target-marking pattern, which could be either ground marking with illuminating flares followed by red and/or green target indicators or sky marking with red and green star flares. The master bomber and the supporting Pathfinders had to drop their markers and repeat markers over an eleven-minute period, while the main force had to watch for the markers, bomb the center of them if they could, and then make for home. The whole combat force typically extended for twenty miles, was six miles wide, and flew in staggered formation, the highest aircraft some 4,000 feet above the lowest. Crews had to fly low over England, then climb to 14,000–15,000 feet, then increase speed and fly at 18,000–20,000 feet for bombing, finally falling away from the target zone to 12,000 feet, back to 18,000 feet for the return flight across Germany, 12,000 feet again at the European coast, and not below 7,000 feet on crossing back over English territory. Then came the debriefing interviews and the post-raid assessments.¹⁹⁵

The Eighth Air Force derived a lot from British experience, but its typical raid preparations and combat showed an even greater awareness of the complexity of the task. The principles guiding target selection dwarfed the simple list of cities and their industrial importance given to Harris. Target selection involved assessing the strategic importance of a target, working out the degree of “cushion” in the economy for substituting or dispersing output, calculating the depth of a target system (how far away a product might be from frontline use), judging the recuperative possibilities of a target, and weighing up its vulnerability and the capacity of the air forces to destroy it by researching its potential structural weaknesses and susceptibility to damage.¹⁹⁶ The material was collected and evaluated by the Enemy Objectives Unit based in the American embassy in London under the leadership of Colonel Richard D’Oyly Hughes, a former British officer who had taken American citizenship in the early 1930s. His team of economists visited British industrial plants to work out the most vulnerable part of each type of target and then applied the knowledge to detailed photographic material on the German equivalent. In the evenings they relaxed from their efforts by working out fruitless statistical teasers—“How many sheep are there in Bavaria?” “What is the most economical land route from Gdansk to Gibraltar?”¹⁹⁷ Their labors proved most effective with assessments of damage to capital-intensive targets such as oil refineries or synthetic oil and rubber producers. This material was taken and put into operationally useful form for air force units.

The typical Eighth Air Force operational procedure reflected a managerial ethos that was quite distinct from British practice. American officers had in many cases been drafted into the air force from business and professional backgrounds, which prepared them for the vocabulary and categories typical of modern managerial practice. The formal procedure laid down in July 1943 reflected that culture: a conference of key personnel at four in the afternoon before the operation at which the prospective weather determined the target to be attacked; target folders checked; fighter escort informed; calculation of type and weight of bombs and number of aircraft; notification of assigned combat groups; finally, determination of axis of attack, rendezvous point, route out, initial point (near the target run-in), altitudes, aiming point, rally point (just outside target area), and route back. The resulting field order was then sent by teletype to the combat units involved.¹⁹⁸ A second procedure then took place at the airbases of the different bombardment wings, with an operational briefing for all commanders and crew for approximately two and a half hours covering the following: plan for formation (general), approximate time for turns, sun position, power settings, intelligence information, and weather prospects. Detailed planning was essential because in daytime bombing the force had to fly in tight formation. The Eighth Air Force flew at 25,000 feet, each wing flying in three staggered combat boxes, covering a height of 3,000 feet to maximize the firepower of the group, before breaking into bombing formation in approach to the target. Separate briefings were then held for pilots and copilots (sixteen items), navigators and bombardiers (six items), gunners and radio operators (three items). Watches were then synchronized. At 8 a.m. on the morning of the operation a decision to go ahead or postpone had to be taken based on the current state of the weather. Commanders had an obligation, according to a manual on tactics, “to work out each mission in minute detail. The struggle here is of the life and death variety.”¹⁹⁹

Somehow or other all the detailed calculations, operating plans, and contingencies had to be mastered and put into effect by the expensively trained crews. There was always to be a gap between the ideal operation laid down by the military bureaucracy that ran the offensives and the reality of combat. Unanticipated factors of all kinds, not least the extent and combativeness of the enemy forces, undermined the best-laid operational procedures. Given the technical sophistication of much of the equipment, the large number of freshman crews to be initiated on each operation, and the vagaries of weather and navigation, it is perhaps surprising that bombing operations achieved as much as they did. Almost all the flight crew were between the ages of eighteen and twenty-five, a large number of them between eighteen and twenty-one; a few who lied about their age flew heavy bombers aged just seventeen. Almost nothing of what they experienced in training could prepare them for what happened by day or by night over Germany. For Bomber Command crews there was extreme cold for much of the time unless they wore layer upon layer of protective clothing; there was a numbing tiredness on operations that could last eight or nine hours, using up the body’s natural adrenaline supplies and requiring chemical stimulants (commonly amphetamines); with the decision taken in the spring of 1942 to have just one pilot, the crew had to hope that one of their number had enough basic flying skills to get them to their target and back if the pilot was killed or incapacitated. There was the constant fear of night fighters and antiaircraft fire or of being coned by searchlights, to which large nighttime formations added the danger of collision or bombs from the invisible aircraft above in the bomber stream.

The Eighth Air Force crews had some advantages; the B-17 Flying Fortress was less cold to fly in, and they were provided with good thermal clothing; each aircraft had a pilot and copilot; attacking aircraft were more easily visible, though the limited range of the B-17 machine guns meant that rocket- and cannon-firing fighters could damage the bombers before facing risk themselves. Other factors were shared. The experience was frightening and the accounts of a great many airmen understandably recall fear as a very primary emotion. “I was scared all the time,” recalled one veteran, “but I was more scared of letting the rest of the crew see.” Aircrew were commonly sick as a result of the long, bumpy flights. Their priority was to complete the mission and return to base. “You bombed a target and got the hell out and got home, there wasn’t much glamour about it,” remembered another. Their primary loyalty was to the other crews about them. “I never worried about the people down below,” said one pilot. “I was more concerned with the ones in the air.” This was perhaps an understandable moral concern. The permanent dangers to which an aircrew was exposed and the sheer mental and physical demands of combat, surrounded at times by dead or dying companions, with jammed guns or engines knocked out, created a temporary nightmare world in which the one hope was that their aircraft and crew would not be next. After the Regensburg raid one American commander who reached Tunisia reported that he and his crew “felt the reaction of men who had not expected to see another sunset.” He recommended reducing the standard tour of thirty operations to twenty-five if crew were not to collapse from the psychological pressure. “Survival was our thing,” concluded one veteran.²⁰⁰

The success of bombing operations depended almost entirely on the quality and training of the crews, but the pressures to which they were subjected placed often insupportable demands on their psychological equilibrium. This situation arose not only because of the natural stresses of combat, but because of the curious social situation in which bomber crews found themselves. Although they were regularly called upon for operations that provided hours of tension and endeavor, once back at base there might be days, sometimes longer, before the next operation. In that interval crews were free to go to the local town, meet girls, reunite with wives or partners, and enjoy a variety of forms of recreation. Cinema attendance at Eighth Air Force bases reached a million a month by November 1943, stage-show attendance 150,000.²⁰¹ This meant that bomber crews had a cycle of relief and anxiety distinct from the emotional pattern of ground combat troops. In the first years of the offensive, survival rates were low, so that life at base was also about reconciling the loss of companions, relishing survival, and anticipating the next operation. Casualties were high not only from combat but because of routine flying accidents. In Bomber Command some 6,000 crew were killed in accidents in 1943–44; the Eighth Air Force suffered in 1943 as many as 8,800 losses in combat and a further 2,000 from noncombat accidents.²⁰² Death or German imprisonment was more common than serious injury, which numbered only 1,315 in the Eighth Air Force up to the end of 1943, mostly to the hands, neck, and head. The American statistical record described those who had finished their tour of thirty operations and returned to the United States as “Happy Warriors”; they were certainly lucky warriors, constituting less than one-fifth of the crews sent to Europe.

The one form of often hidden casualty that bombing encouraged was psychiatric. The stress of combat, or combat fatigue, was not in doubt. Questionnaires from the USAAF Psychological Branch to squadron commanders found that they valued “judgment” and “emotional control” far higher than practical skills among cohorts of incoming pilots.²⁰³ The psychological reaction to flying stress depended partly on the personality of the individual crew member, partly on the nature of the experiences or dangers to which he had been exposed. Most crew on a tour of twenty-five or thirty operations died before they reached their total. Of those who survived, a small proportion became medical casualties as a result of stress, but almost all suffered from some degree of fear-induced anxiety, which was observed to get worse the longer the operational tour lasted. American psychiatrists reported heavy drinking, psychosomatic disorders, and long periods of depression among crew who carried on flying.²⁰⁴ In Bomber Command the tendency was to blame any exaggerated state of anxiety on “lack of moral fibre” (LMF), a stigma designed as an emasculating deterrent to any sign of weakness. Harris thought that among his crews only a quarter were effective bombers, the rest merely there to give German antiaircraft guns something to shoot at.²⁰⁵ Air force medical staff, on the other hand, found that there were very few records of cowardice, despite the popular fear among aircrew that they might be regarded that way if they broke down. In both the RAF and the Eighth Air Force it came to be recognized that regular air operations induced particular forms of neurosis that had little to do with a lack of spirit and everything to do with the harsh experiences of daily flying. Eighth Air Force was instructed to rest and rotate tired crews but to isolate those whose behavior might contaminate the efficiency of their unit. These cases were divided between the categories of “flying fatigue” and “lack of moral fibre.” To the former there was no stigma attached, but the latter were to be removed from flying status, stripped of their commission, and sent home in disgrace.²⁰⁶ Those who developed serious psychoneurotic symptoms were sent to special hospitals to undergo narcosis therapy, and many were subsequently returned to duty, including combat flying. By early 1944 it was found that around 3 percent of flying officers (of those who survived) were removed from flying status before completing a tour of twenty-five operations.²⁰⁷

In Bomber Command the treatment of flying fatigue could be much harsher if unit commanders were prejudiced against the idea of psychiatric casualty.²⁰⁸ But like the Eighth Air Force, a system of classification came into use that allowed the neuropsychiatrists in the RAF medical service to distinguish between those with neurotic conditions capable of diagnosis and possible treatment and those classified as “waverers,” defined as fully fit but fearful. Flying stress was accepted as an understandable reaction to “severe combats, crash landings, ‘bale-outs’ and ‘shaky-dos’ in general.” Those who ended up in front of the psychiatrist were classified in four categories based on a predisposition to neurotic behavior (usually defined by character assessments or family history) and degree of flying stress. Those with a high predisposition or a high level of stress were deemed to be medical casualties and withdrawn from flying without penalty. Those with low predisposition but marked flying stress or those with neither predisposition nor serious evidence of stress were defined as lacking in confidence, and a judgment had to be made about whether they were also guilty of “lack of moral fibre.” It was never easy, as one RAF psychiatrist put it, to tell “whether a man’s inability to continue flying is his fault or his misfortune, whether in fact it is due to simple lack of confidence or of courage, or whether it results from nervous predisposition or illness outside his control.”²⁰⁹

These were fateful decisions for the men involved, since those deemed not to be medical cases (approximately 25 percent of those referred for assessment) were dealt with by an executive board that tended to assume cowardice on the part of the men in front of them. Yet the medical casebooks show that individuals were often subjected to a series of traumatic combat experiences sufficient to challenge the mental stability of even the toughest character:

Flight engineer, 20 raids, 150 operational flying hours: “He had been badly shot up on four occasions. On the last of these, after being attacked by a night-fighter, the port engine of his machine caught fire, the mid-upper gunner was badly injured and the rear gunner was killed. . . . The rear-gunner’s body was burning and motionless. . . . He had to use an axe to hack off bits of the blazing turret and also parts of the rear-gunner’s clothing and body, finally letting the slip-stream blow them all away. . . . The wireless, the hydraulics and the tyres had all gone and a crash-landing was made. . . . 10 days later he was in a nervous state with tremulous hands and sweaty palms. He had some headaches, felt unable to relax, was depressed, preoccupied and unable to concentrate. His appetite was bad. . . . When I asked him how he would sum up his feelings in one word, he said ‘fear.’”²¹⁰

On this occasion the decision was made to allow the officer a spell of noncombat duty, rather than assume cowardice. But there were other cases where fear was imputed by the psychiatrists, often unjustifiably, with the result that an airman could be stripped of his commission and the right to fly. In the RAF, 8,402 aircrew were examined for neurosis from 1942 to 1945, of whom 1,029 were declared LMF, 34 percent of them pilots. The best postwar estimate has suggested that Bomber Command crews provided one-third of those figures, which works out at perhaps 20 per month, a remarkably small proportion of all those regularly exposed to the stress of combat flying.²¹¹ Most medical reports on the air forces indicated high morale despite the high casualty rate. Eaker claimed that morale was not affected so much by losses as by the knowledge that a raid had been ineffective.²¹² Psychiatrists nevertheless found that one of the most important motivations was the desire to make it through to the end of the tour still alive. Since only one in four completed one tour, and one in ten a second one, survival remained a primary drive despite the stresses. Research in the Eighth Air Force found that a large proportion of men returning to the United States after a completed tour of duty displayed “subjective anxiety” symptoms: “weight loss, insomnia, severe operational fatigue, and loss of efficiency.”²¹³

These were the men sent by night and by day against German targets in the context of steadily increasing losses on the major raids. During 1943, Bomber Command lost 15,678 killed or prisoners of war, while the Eighth Air Force lost 9,497, almost all of them in missions against German targets.²¹⁴ The escalating costs of the offensive presented the bomber commanders once again with questions about the strategic value of what they were doing. The reality of tactical stalemate coincided by chance with a revival of the hope, largely abandoned since 1941, that bombing might induce a social or political crisis so severe that it would critically undermine the German war effort. Though Harris saw area bombing chiefly as a form of economic attrition, he never entirely excluded the possibility that his bombing might provoke a political bonus, even to the point of German surrender, and he was happy to fuel such speculation if it strengthened his hand. Air Intelligence, for example, was impressed by his claim that the destruction of half the German urban area would provoke collapse, even if the Gestapo and the SS (Schutzstaffel) were determined to “prevent insurrection.”²¹⁵ For some months political intelligence in Britain, encouraged by German difficulties on the Soviet front and in the Mediterranean, had been suggesting that there might be a positive answer to the question “Will Germany crack?” and that bombing could supply it. In September 1943 the Joint Intelligence Committee prepared a long paper on the “Probabilities of a German Collapse” in which Germany’s situation in the autumn of 1943 was compared with the historical reality of the collapse of the German home front in the autumn of 1918. The JIC thought the conditions of life in the bombed cities much worse than in 1918 and the signs of revolutionary discontent increasingly evident despite the brutal nature of the dictatorship.²¹⁶ In November 1943 an even more optimistic evaluation was produced by the British Political Warfare Executive on the creation in bombed cities of a “new proletariat” with a communist mentality, which might yet create a revolutionary crisis in Germany before the winter was over. In January further intelligence was sent to Churchill on social unrest in Germany which suggested that “the more we bomb, the more satisfactory the effect.” Churchill underlined the sentence with his trademark red pencil.²¹⁷

For Churchill the promise of a German collapse revived the confident assumptions about the political impact of bombing that he had harbored ever since the offensive began in 1940. The evidence in the autumn and winter of 1943 was nevertheless slender, based to some extent on imagining what bombing on such a scale might have meant if it had been British rather than German cities under the hail of bombs. American political intelligence was in general dismissive of the idea that bombing alone could generate a German collapse. Spaatz rejected entirely the value of popular war willingness as a target: “Morale in a totalitarian society is irrelevant so long as the control patterns function effectively.”²¹⁸ American assessments of the revolutionary potential of the German working class focused on the “negative character of its assumption of power” in 1918, following the kaiser’s abdication, and the failure of the German left to stop Hitler. Arnold asked a “Committee of Historians” for their analysis of the prospects for German collapse. The nine historians included distinguished names—Bernadotte Schmitt, Edward Mead Earle, Louis Gottschalk—with experience of writing the history of war and revolution. They concluded that although German morale had deteriorated during 1943, the existence of Nazi control “gives no encouragement to the supposition that any political upheaval can be anticipated in Germany in the near future.” They acknowledged that there was a superficial resemblance to the final days of 1918, but their report concluded that Allied insistence on unconditional surrender, the lack of any effective avenue for popular discontent, and the contrast in the military situation “make the seeming analogy invalid.”²¹⁹The key problem identified by all the critics of the idea that Germany would imminently crack was the exceptional capacity of a totalitarian state to exact obedience. If the German people were “discouraged, disillusioned and bewildered,” as intelligence reports suggested, they still appeared to have a fear of state terror more powerful than the fear of further bombing.²²⁰ “Even when public morale is desperately low,” remarked Portal’s deputy, Norman Bottomley, in a speech in the spring of 1944, “general collapse can for a long while be staved off by a ruthless and desperate party system and a corps of brutal Gestapo hangmen and gangsters.”²²¹

These projections were, as it turned out, broadly correct. The bombing made the German population more rather than less dependent on the state and the party. Like the Blitz, Allied bombing created largely passive responses to the problem of survival. In its monthly news digest in March 1944, American air intelligence published a translated article on the air offensive from the German Berliner Börsen-Zeitung, which seemed to sum up the frustrating reality of an attrition war in the air:

A war with its focal point centred in the air is not the shortest, as was once believed, but on the contrary the longest and most meaningless in its accumulation of destruction . . . particularly as even the greatest terror gradually wears off or corresponding counter-measures are found. Thus the time when it was thought that air offensives alone could force Europe to capitulate and that the Anglo-Americans could then march in with music had disappeared into the dim distance.²²²

The Committee of Historians concluded from their assessment of Allied strategy and German staying power that the defeat of Germany was only possible with continued Soviet pressure from the east, continuous bombing from Britain and Italy, and one or more large-scale invasions of German-occupied western Europe. “It seems clear,” continued the report, “that bombing alone cannot bring about that defeat in the spring of 1944.”²²³ The stalemate in the bombing war could only be reversed by military means. “Our first objective,” wrote General Doolittle to his commanders on assuming control of the Eighth Air Force in January 1944, “is to neutralize the German fighter opposition at the earliest possible moment.”²²⁴