Chapter 3

The “Battle of Germany,” 1944–45

In June 1943, Robert Lovett, assistant secretary of war responsible for the air force, wrote a long memorandum for Arnold in which he analyzed the problems facing the American air offensive and suggested solutions. The most important issue he identified and emphasized was “supply long range fighter protection to help the B-17s.” He suggested designing built-in additional fuel tanks for American fighter aircraft but in the interim adding two wing tanks to the new P-51 Mustang fighter. He concluded, “This is a ‘must.’”¹ The P-51 did not see long-range service over Germany until the spring of 1944, by which time other fighter aircraft converted to longer range were already in service to protect the American bombers part of the way to their targets. But it proved to be a critical explanation for eventual victory in what Major General Frederick Anderson, Spaatz’s deputy from January 1944, called “the battle of Germany.”²

The need for long-range fighters matched the need the German Air Force had experienced in the Battle of Britain in 1940 but had failed to solve. Success in defeating the German Air Force in their own airspace—the “Battle of Germany”—depended on establishing air supremacy, and this in turn relied on the extent to which the Eighth Air Force could use large fighter forces to destroy enemy airpower over Germany itself. Fighter-to-fighter combat and counterforce bombing was the solution not only to the expansion of the bomber offensive but also to the eventual success of Allied invasion in the west and success on other European fronts. What the Eighth (supported by the Fifteenth Air Force flying from Italy) was now engaged in was less a strategic air offensive, more the conduct of a “grand tactical” air battle that resembled in many ways the campaign waged by the German Air Force in 1940. It was belated recognition that even in a modern strategic air war, destruction of the enemy air force and its resources rather than destruction of the enemy home front was the essential condition for eventual victory.

“The Arithmetic of Impending Ruin”

There has been since the war much discussion of why it took so long for the U.S. air forces to develop fighter aircraft with long-range capability able to contest air superiority over Germany. Spaatz had been a witness to the German raids on England in 1940 when the need for fighter cover to protect German daylight bombing had been self-evident. The planners in 1941 who drew up AWPD-1 emphasized that the development of escort fighters that could fly as far as the bombers was “mandatory.”³ Arnold as early as February 1942 had asked for all new fighters to be developed with auxiliary tanks. In June 1943 he ordered a crash program to ensure that full bomber escort could be provided by early 1944.⁴ There was no shortage of high-quality fighter aircraft designed with longer range than the British Spitfire (which had supplied limited fighter escort in 1943). The Lockheed P-38 Lightning, a radical twin-engine, twin-boom fighter, was a prewar development that had long-range extra fuel tanks built into the design. Its development was delayed and it entered service in the summer of 1942; it was used in North Africa as a low-altitude battlefield aircraft, a role that suited it poorly. Late in 1943 two groups were allocated to the Eighth Air Force and at once extended potential escort range as far as Leipzig, though in numbers too small to transform the offensive, and with persistent technical problems with the engine. The mainstay of the Eighth Fighter Command in 1943 was the Republic P-47 Thunderbolt, a high-performance fighter/fighter-bomber designed in 1940 around the Pratt and Whitney R-2800 radial engine. It could carry two external fuel tanks to boost range as far as the German frontier when drop tanks were first installed in July 1943, but little effort went into modifying the P-47 so that it could reach far into Germany. Eaker gave auxiliary tanks a low priority among his many other problems. Yet with larger tanks the P-47 could by the spring of 1944 fly as far as Hamburg, where before it had been confined to an arc that reached little farther than the Low Countries and the German border.⁵

The one aircraft that promised to transform the air war over Germany was the North American P-51 Mustang recommended by Lovett. Originally designed to meet a British requirement in 1940, it began service with the RAF (under the name Apache) in November 1941. British engineers fitted it with a Rolls-Royce Merlin engine and revolutionized its performance, increasing speed, rate of climb, and maneuverability. News reached Washington via the American embassy in London and Arnold immediately saw the aircraft’s potential. By November 1942 he had placed initial orders for 2,200, fitted with Merlin engines made under license in the United States.⁶ Since the P-51 was supposed to fulfill British orders, Arnold had once again to renege on the agreement. After a stormy exchange with Portal in the late autumn of 1943, he got his way.⁷ The P-51 entered service in early December 1943 with drop tanks that could take it 475 miles into Germany; when it finally came onstream in significant numbers in the spring of 1944 its range, with new tanks, could take it farther than Berlin and even as far as Vienna. Most accounts of the battle for air supremacy credit the P-51 with the destruction of German fighter defenses, but rather like accounts of the Battle of Britain, in which the Spitfire has always been privileged over the Hurricane, the sturdy and less glamorous P-47 Thunderbolt bore the brunt of the first months of the Battle of Germany. On the day the P-51 was introduced to combat against targets in France, December 5, 1943, there were 266 P-47s but only thirty-six P-51s. Three months later, on the first deep daylight raid into Germany against Leipzig, there were 688 P-47s and just seventy-three P-51s. By the end of March 1944, the point that some historians have seen as the moment when air superiority passed to the Eighth Air Force, there were still more than twice as many Thunderbolts as Mustangs.⁸

The explanation for the slow evolution of a long-range fighter capability lies not with the technology but with the Eighth Air Force commanders. Eaker had always believed in the self-defending capability of the large daylight bomber formation. The prevailing tactical assumption in operations was “the security of the force”; the larger the bomber stream, the more secure it would be.⁹ The Eighth Fighter Command under Brigadier General Frank Hunter shared Eaker’s view that unescorted bomber operations were possible, and for much of the summer and autumn, when Eighth Bomber Command losses were rising, he ordered fighter sweeps across northern France and the Low Countries that on some occasions encountered no German aircraft at all. When it was insisted that the P-47s escort the bombers more effectively, the range was still too short to provide more than limited assistance, and made shorter still by the order to fly a weaving route next to the bomber formation to match its speed. In August 1943, after the first Schweinfurt raid, Arnold insisted on sacking Hunter and replacing him with Major General William Kepner, a dedicated fighter general, popular with his crews, who saw the role of his command to fly deep into Germany in order to destroy the German fighter force. Eaker opposed the change of commander and remained lukewarm about the effort to use fighters, rather than his bombers, to achieve the air superiority required from the Pointblank offensive. Destroying enemy fighters he saw as “the secondary job”; the primary task was dropping bomb loads as accurately as possible on strategic air force targets.¹⁰

Arnold’s persistent dissatisfaction with the performance of the Eighth Air Force speeded up the decision to activate the Fifteenth Air Force in the Mediterranean to attack Pointblank targets from the south, where the weather was better. Without notifying Eaker, Arnold asked the Combined Chiefs at their meeting on November 18 to approve the reorganization of American air forces by appointing an American strategic air commander for both European theaters, responsible for the Eighth, Fifteenth, and Ninth Air Forces and, if possible, Bomber Command. The Combined Chiefs agreed to the rearrangement on December 4 (with the exception of Bomber Command, which Portal refused to hand over), and Arnold got support from Roosevelt and Churchill.¹¹ Arnold asked Spaatz to return to Britain to take up the post of commanding general, strategic air forces, on January 1, 1944. He brought with him Major General James Doolittle, commander of the Northwest African Strategic Air Force, and a firm advocate of bomber escorts. Spaatz took over Eaker’s headquarters, while Doolittle commanded the Eighth from “Widewing,” up until then headquarters for Anderson’s Bomber Command. Anderson became chief of operations to Spaatz. Arnold had been insensitive enough to notify Eaker of the change in command by telegraph back on December 19, rather than in person, consistent with his testy treatment of Eaker earlier in the summer. Eaker objected vehemently to the change, but his objections were overruled; on January 6, Doolittle took over the Eighth and Eaker left to command strategic forces in the Mediterranean at just the point when large numbers of bombers and escort fighters were at last coming through the pipeline to transform the capability of the air force Eaker was compelled to abandon.¹²

In any assessment of the success in establishing air superiority over Germany, the change in American leadership is clearly central. Spaatz, Doolittle, and Kepner shared a common strategic outlook on the importance of combining the indirect assault on air force production and supplies through bombing with the calculated attrition of the German fighter force through air-to-air combat and fighter sweeps over German soil. Spaatz spent some weeks reviewing the offensive in January 1944 and then told Doolittle that destroying German fighter strength and increasing the tempo of attacks on German aircraft production was “a critical deciding factor in Germany’s defeat.”¹³ Doolittle was from the start eager to use his large force, now with more than 1,000 bombers and 1,200 fighters, to destroy the German air arm. Commenting to Spaatz on the plans for completing the Combined Bomber Offensive, he was critical of the idea of pursuing “economic” bombing, and argued for making attacks on the enemy fighter force in the air and on the ground the “primary consideration,” as it had been when he was a commander in the Mediterranean.¹⁴

Kepner had already begun to transform the tactics of fighter support before Doolittle’s appointment. The key was to allow the fighter escorts to engage the enemy fighter force and not simply protect the bombers; this had been the dilemma facing German fighters in the Battle of Britain, when they were eventually compelled to fly as close support for the bomber stream and lost their combat flexibility. From January 1944 onward, American fighter units were ordered to “pursue the Hun until he was destroyed.”¹⁵ The new tactic of “Free Lance” allocated some fighter planes to abandon the bombers entirely and seek the German force wherever it was to be found. The escort aircraft, flying in loose groups of four, ranged up to seven or eight miles away from the bomber stream in search of combat. On the return flight they were encouraged to fly at a low level to strafe German airfields or attack German fighters taking off or returning to base. To maximize combat time, a system of escort relays was set up in which each stage of a bomber’s flight would be protected by fighter units assigned to a particular stretch, so that they could fly direct to the rendezvous point rather than lose precious fuel flying slowly with the bombers. P-47s guarded the first and last legs of the route, P-38s the intermediate stretch, and the very long-range P-51s the area close to the target zone. The success of the change in tactics depended first on a much-enlarged supply of fighter aircraft and pilots, with improved levels of maintenance, and the exploitation of the RAF “Y” radio-intercept service, which made it possible for American fighters to be directed to the point where German aircraft were themselves assembling in formation.¹⁶ The object was to leave the German enemy no respite from the threat of combat and to impose an insupportable level of attrition by deploying more fighter aircraft than the enemy. “The arithmetic in itself,” Anderson told Arnold in February 1944, “spells impending ruin.”¹⁷

The German Air Force did not remain passive in the face of the growing American threat. The driving force behind the reorganization of air defense and the expansion of fighter output was Göring’s deputy, Erhard Milch, who understood more clearly than his master that “the homeland is more important than the front.”¹⁸ The allocation of priority to the defense of the Reich and to fighter production brought about not only a regular process of tactical and technical readjustment, but a major change in command and organization as well. In August 1943 the chief of staff, Hans Jeschonnek, who effectively carried the weight of high command in Göring’s increasing absence, found the constant criticism and abuse from his commander in chief over the bombing offensive impossible to withstand. On August 19 he shot himself, leaving behind two letters for Hitler’s air adjutant condemning Göring’s incompetent leadership. Jeschonnek was not entirely blameless, since he had continually emphasized the importance of airpower at the fighting front rather than defense of the home territory.¹⁹ He was replaced by Colonel General Günther Korten, whose relationship with Göring was better, but unlike Jeschonnek, he was committed to the idea of strengthening home air defenses and had Hitler’s support for doing so. In November, Kammhuber was removed from his post, one of the remaining obstacles to reorganizing the defensive system. In northern Germany, Fighter Corps I (Jagdkorps I), responsible for the fighter defense of most of western and central Germany, was expanded and placed under the command of Lieutenant General Josef “Beppo” Schmid, best known for supplying overoptimistic intelligence during the Battle of Britain. From a single fighter wing in January 1943, Schmid’s new command had eleven wings and twenty fighter groups by the end of the year. In December 1943, Hubert Weise, Luftwaffenbefehlshaber Mitte, was replaced by Colonel General Hans-Jürgen Stumpff, the former commander of Air Fleet 5; on January 27 the command was renamed the Reich Air Fleet (Luftflotte Reich), now responsible for the coordinated control of the entire defensive air war against the Allied bombers.²⁰ The process of creating a single centralized air defense of Germany was completed in February 1944 with the transfer of antiaircraft artillery and the German air-warning system to direct control by the Reich Air Fleet and the local fighter divisions. The system now more closely resembled the centralized control structure set up by Fighter Command in Britain in 1940. From Stumpff’s headquarters in Berlin it was possible, using the radar information from Fighter Corps I, to communicate a running account of the air battle to the fighter units to ensure a concentrated response; for his part, Schmid had no fewer than 148 direct telephone lines to fighter stations and control centers.²¹

The German Air Force knew a remarkable amount about the British and American air forces. Most of the information came from downed Allied aircraft and interrogated prisoners. The bombers’ predictable tactics and long flying time over German territory in the last months of 1943 had contributed to the escalating loss rates imposed on each bombing mission.²² The electronic war, which had swung briefly in the Allies’ favor with the use of Window over Hamburg, was more evenly balanced by the end of the year. German researchers quickly discovered ways to neutralize the effects of Window with two devices, Würzlaus and Nürnberg, which allowed the more skillful radar operators to distinguish between Window echoes and an airplane; by the end of the year, 1,500 Würzburg radar devices had been modified. The German Telefunken researchers came up with a new air radar device, code-named SN-2, that could operate impervious to Window interference, and a crash production program was begun. The new Allied H2S radar navigation could also be tracked by the end of 1943 using a new homing device, Naxos-Z, which enabled German night fighters to track the RAF Pathfinder Force; it also proved possible to get a bearing on the Allied bombers that were not carrying H2S by using their Identification Friend-or-Foe mechanism. Both breakthroughs contributed to Bomber Command’s escalating losses. The Eighth Air Force began to use Window (code-named Chaff in the United States) on December 20, 1943, at the same time as introducing a Würzburg jammer known as Carpet to reduce losses by radar-guided antiaircraft fire. Here again German radio engineers found a partial solution by introducing a modification known as Wismar, which allowed the radar to switch frequencies and avoid the effects of Carpet, though by this time the tactical battle between the two air forces had rendered electronic protection less important.²³

The keys to German air defense were assumed to be production and manpower. To meet the threat of daylight bombing, the antiaircraft artillery was substantially increased in early 1944, with 1,508 heavy batteries (5,325 guns), 623 light batteries (9,359 guns), and 375 searchlight batteries (5,000 lights of 200- or 150-centimeter diameter). Output of antiaircraft guns reached a peak in 1944 of 8,402 heavy and 50,917 light guns, but the wastage rate of barrels doubled over 1943 because of the increased bomber activity.²⁴ An additional 250,000 personnel had to be found in 1944, mainly recruited from Soviet prisoners of war, Italian volunteers from Mussolini’s Italian Social Republic, air force wounded, and young German volunteers. This represented a damaging dilution of the quality of antiaircraft personnel. By the spring of 1944 some 111,000 women also served in the German antiaircraft defense system. To navigate the regime’s confused stance on employing women, posters reminded the female volunteers, “The woman in a soldier’s post but still a woman!”²⁵ The antiaircraft batteries by 1944 were organized increasingly in large groups of heavy guns—Grossbatterien made up of three regular batteries—to produce more concentrated fire, but this made heavy demands on a less skilled and less robust workforce. Yet over the course of the year antiaircraft artillery came to replace the fighters as the main means for destroying or damaging enemy aircraft; the major industrial targets were protected by defensive strongpoints of no fewer than three Grossbatterien.²⁶ Throughout the campaigns of 1943 and 1944, antiaircraft damage to Allied aircraft was extensive. An American raid on Berlin on March 6, 1944, resulted in damage to 48 percent of the 672 bombers that reached the target. Only faltering supplies of ammunition prevented antiaircraft fire from being more effective.²⁷

The accelerated production of fighter aircraft also faced problems in late 1943, partly because aircraft production was still controlled by the Air Ministry while the rest of the armaments economy had been centralized under Albert Speer’s Ministry of Armaments and War Production, and partly because of Göring’s renewed efforts to revive German offensive airpower by switching resources to bombers again in the winter of 1943. The current plans for producing more than 30,000 fighters in 1944 and 48,000 in 1945, drawn up by Milch’s planning staff in the German Air Ministry, also lacked realism, not least because of the problem of fuel supply. Yet the figures matched what the crisis in the skies over Germany seemed to require.²⁸ Milch collaborated closely with Speer and the head of his technical office, Karl-Otto Saur, to reduce the different models of each aircraft type—eventually reducing the models from forty-two to five—and to speed up dispersal programs. But the problems posed by Göring’s revival of bomber plans pushed Milch, for political as well as practical reasons, to offer control over aircraft production to Speer to achieve a long-overdue rationalization of the whole production structure.²⁹ In February the two men reached an agreement to run together an emergency “Fighter Staff” (Jägerstab) with Saur as its director, and it was established with Hitler’s agreement on March 1, 1944. As a result, in 1944 three times as many fighters were produced than in 1943, in the hope that this would be sufficient to hold back the Allied bombers long enough to allow the whole German aircraft program to revive and expand.³⁰

It was nevertheless evident by the end of 1943 that sheer numbers of German fighter aircraft were not the entire solution. The production of aircraft had to be balanced against losses, and despite the success rate of German day and night fighters against the major raids of the autumn and winter, the cumulative attrition of the fighter force made it difficult to expand overall force strength despite the very substantial increases in output. Although 3,700 day and night fighters were produced between September and December 1943, the force at Stumpff’s disposal when he assumed command in December numbered just 774 day fighters and 381 night fighters, with serviceability levels of 60–70 percent because of shortages of spares and skilled ground personnel.³¹ This paradox can be explained in a number of ways. Fighter aircraft were compelled to fight in poor weather conditions against bombers now using blind-flying techniques. Commanders sent aircraft out in dangerous conditions (though not usually in fog or heavy cloud), with the result that the accident rate rose sharply again. Icing and misting of the cockpit windows was a particular hazard. Between September and December 1943 the German fighter force lost 967 aircraft in combat, principally with the American P-47 Thunderbolt, but a further 1,052 to accidents.³² The second factor was pilot strength and quality. The high loss rates could not easily be made good by the flying schools, which were under intense pressure to supply crew to every combat theater. The result was a sharp reduction in the length of time devoted to training, which was exacerbated by the careful use of fuel. The hours devoted to training for a new German fighter pilot fell from 210 in 1942 to 112 by 1944; operational training was reduced from 50 hours to 20, and crews could be sent to squadrons with only a few hours’ training on the frontline aircraft they were to fly in combat. Pilots who returned from combat on the Eastern Front found it difficult to adjust to dogfighting with skilled opponents, while pilots drafted in from other branches of the air force, or from air ferrying, were not the equal of enemy crew who enjoyed dedicated fighter training in an entirely bomb-free environment.³³ The result was that by early 1944 the German fighter force was obtaining an average net gain every month of only twenty-six new pilots. The stalemate inflicted on the bomber forces in the autumn created the illusion of German success. In reality the German Air Force was a brittle shield.

The declining skills and rising losses of the German day-fighter force were magnified by the insistence that the object for the force as a whole was to destroy the enemy bomber. This, too, had been a problem for RAF Fighter Command in 1940, when the choice had to be made between stopping the German bombers or fighting their intruding fighter force. German Air Force tactics worked effectively as long as their fighters could seek combat in areas where the bomber force was unescorted. The introduction of longer escort runs in late 1943 transformed the battlefield, though the German Air Force was slow to adapt to the changed reality. Göring famously insisted that the first long-range American fighters to crash near Aachen must have drifted there with the prevailing wind.³⁴ The existing German fighter force was divided between the Me110/Me410 “destroyer” aircraft, armed with rockets and cannon against the Allied bombers, and the more versatile Me109 and Fw190 fighters that were responsible when necessary for air-to-air combat with enemy fighters. Once American escorts appeared, the slower twin-engine German “destroyers” were sitting ducks. The first reaction was to move defense units farther into Germany in the hope that Allied escort fighters would still have a limit to their range. But the heavier destroyers now had to be escorted by the single-engine fighters, which meant that they too would be tied to a role in which they would be at a persistent disadvantage. In March the destroyers were finally withdrawn altogether after one wing of forty-three aircraft lost twenty-six in one raid, but the prevailing German view was still that their single-engine fighters had to try to get close to the enemy bombers to inflict damage, leaving those fighters easier prey to the increasingly aggressive Americans.³⁵ The more flexible the tactics of the Eighth Fighter Command became, the more inflexible the tactical demands on the German Air Force.

These weaknesses were cruelly exposed when Spaatz unleashed his campaign for air superiority over Germany. The eventual success of this campaign could not be taken for granted, not because of the German enemy but because of arguments over strategy among the Allies. There was no question that undermining the German Air Force was now a top priority. But Spaatz had to achieve Pointblank in competition with the demands for the “Crossbow” operation authorized by Allied leaders in late 1943 against the V-weapon silos and installations, and the early onset of bombing tactical targets in support of Operation Overlord, which was expected in February 1944 to absorb at least three months’ bombing effort by the strategic air forces.³⁶ The tension between pursuing Pointblank targets in Germany and the diversion to targets in occupied Europe more directly related to invasion was evident to Spaatz and his commanders. It resulted in prolonged arguments over target priorities, which were finally resolved at a meeting between Eisenhower and the senior Allied commanders in Europe on March 25, 1944, in favor of the “Transportation Plan” for interrupting German rail traffic in northwest Europe. Spaatz was able to start his assault on the German Air Force before these arguments had been properly formulated and resolved—and in the event between January and May 1944 the Eighth and Ninth air forces based in Britain dropped 111,546 (75 percent) tons of bombs on strategic targets against 38,119 (25 percent) on tactical ones.³⁷ The real problem for Spaatz was the difficulty in persuading Harris to share in the task of defeating the German Air Force.

Harris was determined in early 1944 not to abandon the city attacks for a more concentrated assault on German Air Force targets. In January his command was asked directly to abandon indiscriminate area attacks (Harris scrawled “never has been” in the margin of the memorandum) in favor of raids on ball-bearing factories and fighter output as a contribution to the Eighth Air Force effort to establish “free deployment” for the day campaign over Germany.³⁸ Figures were produced by Bomber Command intelligence to show that over one-third of German man-hours had been lost in the bombed cities. Harris told the Air Ministry in early March 1944 that if his force stopped city bombing, German industry would quickly recover and nullify all the efforts his force had made over the previous year.³⁹ When the air minister, Sir Archibald Sinclair, asked Portal for the opinion of the air staff on Harris’s strategy, Portal replied candidly enough that the effort to calculate when Germany might collapse under a certain weight of bombs was “little more than a waste of time”; the air staff, he continued, preferred a strategy of isolating and attacking the vulnerable points in the German structure, whereas Harris just believed in “piling the maximum on the whole structure.”⁴⁰ Portal nevertheless made little effort to get Harris to comply with the American plan to hit air force targets, until pressured to do so by Sydney Bufton, director of bomber operations. Harris was finally ordered to bomb Schweinfurt by a special directive, and Bomber Command obliged on February 24–25. Target marking was generally poor and the damage to the city and its ball-bearing industry “nominal”; only twenty-two bombs fell within the city boundaries, the rest in open country. In this sense Harris’s fear that his force could not hit a small urban target effectively was right.⁴¹

Further raids were made to support the American campaign against Leipzig, Augsburg, and Stuttgart, where there were aircraft and component firms, but the raid on Leipzig missed the Erla aircraft works entirely at a cost of 11 percent of the attacking force, while the raid on Augsburg did little industrial damage but burnt out the whole medieval center of the city. The raids on Stuttgart, mainly through cloud, were scattered, though a lucky hit was made on the Bosch magneto plant. Throughout the period when Spaatz was attacking the German Air Force, Harris persisted in continuing the Battle of Berlin, where losses remained high and the impact limited. An assessment of the attacks on the capital between November 1943 and February 1944 by RE8 showed that only 5 percent of residential buildings and 5 percent of industrial plants had been damaged in heavy raiding.⁴²The attacks made in March on Berlin still brought loss rates of between 5 and 9 percent of the force on each raid. The last major British raid of the war on Berlin, on March 24–25, experienced high winds and resulted in scattered bombing across 126 villages and townships. Some 72 aircraft were shot down, 8.9 percent of the force. In April the final city raid against Nuremberg before the switch to the Overlord campaign showed the persistent limitations of area bombing. A total of 95 aircraft were lost out of the 795 dispatched, the highest loss rate of the war, 11.9 percent. At least 120 aircraft bombed Schweinfurt by mistake, but missed the main area of the city; the remainder bombed a wide area of the German countryside north of Nuremberg, killing sixty-nine villagers. Harris at last recognized that the effectiveness of the German night defenses, as he told the Air Ministry, might soon create a situation in which loss rates “could not in the end be sustained.”⁴³ Between November 1943 and March 1944, Bomber Command lost 1,128 aircraft for little evident strategic gain. Losses among the expanding German night-fighter force were also high, but by the spring they could see that they were gaining as close to a victory as air war would allow.⁴⁴

In the end the defeat of the German Air Force was an American achievement. Spaatz divided the campaign into three elements: Operation Argument to undermine German aircraft production; a follow-up campaign against the German oil industry to starve the air force of its most precious resource; and finally, continuous counterforce attacks against German fighters and their organization. The attack on the aircraft industry, which came to be known as Big Week, was postponed regularly through late January and early February 1944 by poor weather. Attacks were carried out against targets in France and a few deeper raids into Germany, but cloud and snow kept German fighters grounded and increased the risk of accident to American aircrews. On February 19 the weather finally cleared, and for the week until February 26 the Eighth Air Force flew 6,200 sorties against eighteen aircraft assembly plants and two ball-bearing factories. The raids on the first day, February 20, divided the bombers between twelve major targets in Rostock, Brunswick, Leipzig, and half a dozen other smaller towns. The losses totaled only 15 bombers from the 880 that attacked—a rate of only 1.7 percent—and 4 fighters. Losses climbed as the German Air Force grasped the pattern of attacks, and the cost during the week was eventually 158 for the Eighth Air Force (imposed when, for some reason, escorting lapsed) and 89 for the Fifteenth, which attacked from Italy entirely without escort. Only 28 American fighters were lost from the large numbers dispatched on each raid, but the German Air Force lost one-third of its single-engine fighters during February and almost one-fifth of its fighter crew. By contrast, the number of P-51 Mustang fighters available was 90 percent higher at the end of Big Week than it had been at the beginning.⁴⁵

The damage sustained by the German aircraft industry was difficult for the Allies to gauge, not least because air intelligence estimates of German production by this stage of the war greatly understated the reality. The MEW estimate of German fighter production for the first half of 1944 was 655 a month, whereas the reality was 1,581 and rising steadily.⁴⁶ The aero-engine industry, more difficult to disperse and more vulnerable, was not attacked, a failure that Göring later pointed out to his postwar interrogators.⁴⁷ The attacks accelerated the further dispersal of the industry and prompted a program for underground construction in which aircraft had a priority, a planned 48 million square meters of floor space out of a provisional total of 93 million.⁴⁸ Output nevertheless continued to increase rapidly despite the bombing, and this has encouraged the view that Operation Argument effectively failed. The figures show, however, that the Allied attacks, which continued intermittently thereafter, did reduce planned fighter output substantially below expectations. Between January and June 1944, 9,255 German single-engine fighters were produced instead of the planned 12,667, a shortfall of 27 percent. The heaviest loss was experienced in February 1944 with a shortfall of 38.5 percent of planned output.⁴⁹ Not all of this loss was due to bombing, since many other factors affected industrial performance by 1944, but the impact in February almost certainly was. The problem for Allied calculations was the failure to apprehend the rapid conversion in Germany to fighter priority and the successful rationalization and reorganization of aircraft production.

Spaatz also planned to attack oil facilities, particularly those producing aviation fuel, which were more vulnerable than aircraft assembly halls because of the large capital plant involved and the difficulty of dispersing or concealing them. Intelligence on German oil supplies was the reverse of aircraft production, consistently overestimating German synthetic production and imports. Reluctance to renew an oil offensive after the RAF failures of 1940 and 1941 was based partly on the belief that Germany had large concealed stocks available. By the spring of 1944, however, Allied intelligence indicated a growing oil vulnerability in Germany. Spaatz set up a planning committee in February 1944 composed of members of the Enemy Objectives Unit to report on other target systems that would accelerate German Air Force decline, and the committee report, presented to him on March 5, highlighted oil as the principal factor, followed by rubber and bomber production. The economists calculated that enough damage could be done to current oil production to force the German armed forces to consume remaining stocks and that this was the quickest way to undermine fighting power.⁵⁰ Spaatz willingly accepted the argument and used the new oil plan to make his case, unsuccessfully, against the diversion of his resources to the tactical Transportation Plan. The aim to destroy or immobilize twenty-seven key oil targets was presented to Portal and Eisenhower as a surer way to undermine German military mobility at the front line, but the estimate that it might take three months to do so made oil plants, in Portal’s view, a long-term objective. Instead, the Transportation Plan won the day.

In the end Spaatz succeeded in undertaking attacks on German oil targets by sleight of hand. In April 1944 the Fifteenth Air Force began a number of raids against the Romanian oil-producing city of , nominally against “marshaling yards.” In fact the raids hit the oilfield, as intended, and in early May, Eaker gave tacit approval for further attacks on Romanian oil production. Spaatz managed to persuade Eisenhower that German Air Force dependence on oil made it effectively a Pointblank target too and got a verbal assurance that on days when he was not attacking French targets, he could attack synthetic oil production.⁵¹ On May 12, Spaatz finally sent 886 bombers escorted by 735 fighters to attack six major oil plants across Germany. The force lost 46 bombers (32 of them from a bomber division whose escort failed to rendezvous correctly), but the swarms of American fighters destroyed 65 enemy aircraft for the loss of just 7 planes. The high-level Ultra intelligence, produced at Britain’s code and cipher center at Bletchley Park from intercepted German messages, revealed the following day an urgent German order to move all available antiaircraft artillery to protect the synthetic oil plants, including guns that until then had been guarding the aircraft industry. The next raid on May 28 was even more devastating, temporarily destroying output at the oil plants at Leuna and Pölitz in eastern Germany. Spaatz was proved right: the oil targets not only encouraged fierce defense by the German fighter force, but quickly proved debilitating to German forces reliant on a shrinking supply of fuel. Production of aviation fuel was 180,000 tons in March, but had fallen to 54,000 tons in June. So successful were the first attacks that on June 4, two days before the invasion of France, Eisenhower gave formal approval for the oil offensive.⁵²

All the while, Spaatz was driving the Eighth Air Force to impose insupportable levels of attrition on the enemy fighter force. When there were no bomber raids, Kepner was encouraged to send his long-range fighters in wide sweeps over German territory, attacking German airbases and seeking opportunities for combat. For bombing operations Spaatz chose long-distance targets that would compel German fighters to attack the bombers. In March he launched a number of major raids against aircraft production in Berlin, briefly overlapping with the battle Harris had been waging since November. The raids were among the costliest of the Pointblank campaign. On March 6, 730 bombers and 801 fighters left for the first raid on the capital. Fierce battles erupted over the city so that not only was the bombing inaccurate but the raid cost the Allies 75 bombers, though only 11 escorts were lost for the destruction of 43 German fighters. Raids continued throughout March and April, culminating in a final assault on Berlin on April 29 in which the bombing was ineffective and 63 bombers were lost. The German Air Force had reacted to the advent of the long-range escort fighter by creating large concentrations of up to 150 fighters—the “Big Wing” used in the last stages of the Battle of Britain—that were designed to batter their way through to the bomber stream, or, when opportunity presented, to focus entirely on bombers whose escort had failed to materialize. The results for both sides were the highest losses of the war. In April the Eighth Air Force lost 422 heavy bombers, 25 percent of the total force; the German fighter force lost 43 percent of its strength in the same month.⁵³

The arena of daylight air combat over Germany was among the harshest of the air war. American commanders expected a great deal of their crews. “Greater risks are justified,” wrote Anderson to Arnold, “and high losses are to be expected.”⁵⁴ They were able in the end to accept high losses only because a generous spring tide of aircraft and crew was now flowing across the Atlantic. For the German fighter force, high losses made it difficult to keep more than 500 serviceable fighters in the Reich Air Fleet at any one time. The result was that in air-to-air combat, fighter to fighter, the German force was completely outnumbered and the concentrations easily broken up. “An enormous number of us arrived, a crowd of 30, 50, sometimes 60 aircraft,” a captured German fighter wing commander explained, “but each pilot simply attacked wildly at random. Result: each of them was shot down wildly at random.”⁵⁵ The same officer described the decline in German pilot morale over the spring of 1944 as the order persisted to attack only the bombers, when their instinct was to protect themselves by engaging the enemy fighters first. One of Germany’s surviving pilots, Heinz Knocke, later published a vivid diary account of what air combat was like for German pilots in the spring of 1944:

During the ensuing dogfight with the Thunderbolts my tail-plane was shot full of holes, and my engine and left wing were badly hit also. It is all I can do to limp home to our field. . . . Immediately I order a reserve aircraft to be prepared for me to take off on a third mission. It is destroyed during a low-level strafing attack. Two of the mechanics are seriously wounded. No. 4 flight places one of its aircraft at my disposal. . . . When we attempt to attack a formation of Liberators over Lüneberg Heath, we are taken by surprise by approximately forty Thunderbolts. In the ensuing dogfight our two wingmen are both shot down. After a wild chase right down to ground level the Commanding Officer and I finally escape with great difficulty.

Knocke sat in the crew room that evening with the one remaining pilot from his squadron.⁵⁶ Declining morale was not difficult to explain with a one in two chance of surviving, repeated sorties each day, regular and unpredictable low-level attacks, irregular supplies, and little chance of leave. Missions for German pilots became all but suicidal by the time of the Normandy invasion, when hundreds of fighter aircraft were sent west from Germany against odds even greater than the ones they had met in the spring.

For American aircrews the situation was less rosy than German accounts might suggest. Morale dropped for them too during the spring offensive, partly because of high losses, partly because of the demands made on the crews from bad-weather flying. In March and April 1944, eighty-nine bomber crews chose to fly to Swiss or Swedish bases for internment. Conditions were made worse by the decision to abandon automatic repatriation of crews to the United States after twenty-five missions in order to keep up the number of experienced aircrew available.⁵⁷ German interrogation reports of crashed American aircrew found a deep fear of antiaircraft fire, and a strong dislike of the order to conduct low-level attacks against German airfields because of light flak and the tactic of stringing thick steel hawsers (Drahtsperre) across narrow valley approaches to slice into an attacking fighter.⁵⁸ A major hazard was the return flight with battle damage and the difficulty of landing away from base. The crew of one B-24 Liberator bomber, hit by antiaircraft fire over Brussels, bailed out over Kent at the last moment before the damaged aircraft exploded: “I broke an ankle and incurred internal injuries,” recalled the pilot. “The navigator hit a tree and broke his back. . . . The flight engineer had a scalp injury from hitting his head on a rock. In all, we were pretty lucky.”⁵⁹ The high casualty rate made it difficult for American aircrews to form any sense of whether they were winning the battle or not. In the period from January to the end of May 1944 the Eighth and Fifteenth air forces lost 2,605 bombers. Between March and May the American fighter forces lost 1,045 aircraft over Germany and France.

Success only gradually became evident in May and June when Allied bomber losses suddenly fell sharply from the peak in April. By the summer the percentage of attacking bombers actually hit by enemy fighters fell from 3.7 percent in March and April to only 0.4 percent in July and August.⁶⁰The reason can be found in the corresponding German statistics. Between January and June, German aircraft losses on all fronts equaled 137 percent of established strength, 6,259 lost in combat, 3,608 due to accidents, predominantly due to poor weather or pilot error. Despite fighting much of the time over German territory, the German Air Force also lost 2,262 pilots. Most of the losses occurred in Germany or on the Western Front in France and the Low Countries. In June 1944 losses totaled 3,534, only slightly less than the 3,626 aircraft of all types produced that month.⁶¹ This was an insupportable attrition cycle of both German matériel and manpower: even with the increases in fighter output that peaked later in the year, new production was sucked into a whirlpool of rapid destruction. Fighter pilots waited for the Me262 jet fighter, which first flew in March, in the hope that, produced in volume, it might turn the tide.

The point at which Allied air supremacy was established in German airspace is difficult to establish because of the continual, fluid, and incoherent nature of air combat. Some historians date it from the first attrition battles in March 1944, others from the early attacks on oil installations. The head of the Historical Section of the German Air Force, Major General Hans-Detlef von Rohden, argued in a postwar assessment that Allied air supremacy over Germany had been achieved by the time of the Normandy invasion: “Germany had lost the struggle for Air Control.”⁶² A Joint Intelligence Committee evaluation in August 1944 concluded that the German Air Force “can no longer affect the military situation on any front,” which was not entirely true, but did reflect the exceptional degree of operational flexibility now available to American and, increasingly, British aircraft over Germany.⁶³ No date is entirely satisfactory, but by June, when German reserves were sucked into the aerial maelstrom in France, the attrition cycle was, for the moment, complete. This was a situation the German Air Force wanted to reverse. In September 1944 a staff paper reflected on the lessons of the Battle of Britain: “We must try to achieve what England achieved in 1940.”⁶⁴ The larger question posed by the “Battle of Germany” is why the German Air Force failed where in 1940 the RAF—by a narrow margin—succeeded.

There are certainly grounds for comparison. The German Air Force had a substantial fighter force with technology at the cutting edge, particularly after the Me109 fighter was refitted with the more powerful Daimler-Benz 605A engine; aircraft production was concentrated on an emergency fighter program; a large pool of more than 2,000 fighter pilots was regularly available; there was a complex advance-warning system based on sophisticated radar equipment; and the organizational reforms during the winter of 1943–44 created a central control and communications system not unlike the centralized structure available to Hugh Dowding in 1940. The German Air Force had good intelligence warning of incoming attacks and a thorough understanding of enemy tactical weaknesses. Like Fighter Command, the men who fought in the German fighter force were defending their homeland and prepared to take high losses in doing so. As in the Battle of Britain, the German Air Force leaders believed that success in the air was at that point “the most decisive precondition for victory.”⁶⁵

The comparison is nevertheless a superficial one. Germany’s strategic position in 1944 was very different from Britain’s in 1940, fighting as it was on two major fronts in the Soviet Union and Italy and facing growing resistance in other areas of German-occupied Europe. The German priority was not simply to frustrate the Allied search for air superiority but to try to defend a fortress area in central Europe against overwhelming material superiority on all fronts. The strategic crisis explains the emergency program of fighter production, like the British crisis of the summer of 1940. German fighter output reached its wartime peak between the last months of 1943 and the autumn of 1944, though this was achieved in an environment of heavy and continual bombing. As a result, the gap between German fighter production and Anglo-American fighter output (produced in an almost entirely bomb-free environment) was not as significant as the gap in economic resources might suggest. British and American fighter output between January and June 1944 was 11,817; German production over the same six months was 9,489.⁶⁶ In both cases this production was spread among a number of fighting fronts. Yet the Eighth Fighter Command had more than twice the number of fighters available when compared with the Reich Air Fleet, as well as additional support from RAF Fighter Command and the Ninth Air Force. In May 1944 the Reich Air Fleet had 437 serviceable fighters, the Eighth Air Force 1,174. The explanation lies partly in the difference between the two training regimes already noted, which put novice German pilots at a permanent disadvantage. There was also a major contrast in serviceability rates, which were higher for Allied aircraft once the American logistical system was working effectively. Under constant air attack and a manpower shortage, the production and distribution of spares and the supply of adequate ground engineering staff all declined in Germany. More than 9,000 German aircraft in 1944 were lost in transit to Allied air attack before they reached the combat squadrons.

These contrasts were reflected in rates of operational readiness and rates of loss. During the Battle of Britain the peak loss rate for Fighter Command reached 25 percent in September 1940. German Air Force monthly fighter losses were already 30 percent of the force in January 1944 and more than 50 percent by May (see table 3.1). Numerical inferiority was then compounded with the demand that German fighters seek out the Allied bombers rather than fighters, which made them more vulnerable at the moment of attack, and by the decision to assemble large numbers of fighters together (like Douglas Bader’s “Big Wings” in the Battle of Britain); this meant time lost in flight to assembly points and, for pilots who had flown on the Eastern Front in pairs or loose groups of four, a difficult adjustment to flying in larger formations.⁶⁷ The RAF in 1940 avoided both of these operational handicaps by using Spitfires against enemy fighters, Hurricanes against the bombers, while Dowding judiciously resisted the switch to “Big Wings.” The difference between the two sides was not simply a product of economic resources, as is usually argued, but stemmed from operational and tactical choices that rested in the end with those in command.

Table 3.1: Comparative Fighter Statistics, German Air Force and Eighth/Ninth U.S. Air Forces, January–June 1944

Source: Calculated from Richard G. Davis, Carl A. Spaatz and the Air War in Europe (Washington, DC: Center for Air Force History, 1993), App. 9, 22–24; Horst Boog, Gerhard Krebs, and Detlef Vogel, eds., Das Deutsche Reich und der Zweite Weltkrieg. Band 7: Das Deutsche Reich in der Defensive (Stuttgart: DVA, 2001), 105.

German Air Force commanders were also quick to point out to their interrogators after the war that the principal problem they faced was leadership. This is more difficult to assess, and air force commanders were scarcely without prejudice. Most accounts blamed the air force commander in chief, Hermann Göring. Minutes of the regular meetings in late 1943 and early 1944 show a commander full of irate bluster and frustration, prone to impulsive gestures and trite solutions. Though he was capable of sudden bursts of activity, his subordinates found him a bizarre, sadly comical figure. A paratroop general, secretly tape-recorded in captivity in October 1944, entertained his fellow officers with a description of a recent interview with Göring at Karinhall: “There stood the figure and I thought: is it Nero II or a Chinese mandarin? [laughter]. . . . A cloud of all the perfumes of the orient and occident met you half-way exuding over the fat cheeks . . . [All: laughing helplessly].”⁶⁸ Göring, however, blamed Hitler: “You had a great ally in your aerial warfare—the Führer,” he told interrogators in June 1945.⁶⁹ From 1942 onward, and particularly after the failure of the air force to supply the encircled forces at Stalingrad, few major decisions in the air war could be taken without Hitler’s approval or intervention. Yet Hitler did many things right in relation to the air war: he did not in the end obstruct the shift to fighter priority, favored a heavy antiaircraft defense, authorized the dispersal of industry underground, and bullied the air force into prioritizing improvements in electronic warfare.

Göring aside, the other leadership problems stemmed from lower down the air force tree. The coordinated aerial defense of German territory in 1944 fell to an organization that for years had been accustomed to conducting operational air warfare at the fighting fronts. The shift of three-quarters of the fighter force to Germany and the sharp decline in the bomber arm forced a rapid adjustment to an unfamiliar air environment. Stumpff, Schmid, and Korten were relatively inexperienced for the kind of contest they fought in the Battle of Germany. Spaatz, Kepner, and Doolittle had solid experience with just the kind of battle they faced in 1944 and suffered little direct interference from Arnold or Roosevelt in Washington. Although Arnold was also capable at times of irate bluster, he quickly grasped key technical and organizational issues—the importance of the P-51 fighter, the absolute priority for extra fuel tanks, the critical role of logistics—which made his style of management more effective than Göring’s was in 1944 or had been in 1940.⁷⁰ One factor did link the Battle of Britain and the Battle of Germany: the German Air Force did not admit that they had lost either one. In the same document that reflected on how the German Air Force should emulate Fighter Command in 1940, optimistic plans were sketched out for a possible revival of effective fighter defense and a renewed bombing effort, despite the profound crisis now facing German airpower: “The war can only be brought to a satisfactory conclusion if we take the offensive.”⁷¹

Releasing the Hurricane, September 1944–May 1945

The combined offensive was formally reactivated in September 1944 after three months in which the priority for Allied air forces had been supporting the invasion of France and the defeat of German armies in the west. Eisenhower eventually relinquished control of the strategic air forces on September 14, 1944, though he retained the right to request help with the land war when needed. Both bomber commanders were eager to return to what they saw as their primary mission. Spaatz reported to Arnold on the revival of the preinvasion command structure with the comment that “the Hun has still got a lot of fight left in him . . . we must concentrate to kill him off.”⁷² By July, Harris was impatient to restart full-scale bombing because he expected Germany to have recovered fully in five bomb-free months. In August, Portal warned Churchill that there were evident signs of German revival, which would have to be snuffed out by a bombing policy of “continuous attrition.”⁷³

What followed from September 1944 to May 1945 was an Allied campaign with the heaviest weight of bombs and the highest level of German casualties of the war. Over the eight months until German surrender the Eighth and Fifteenth air forces together with Bomber Command dropped three-quarters of the wartime bomb total against a deteriorating German defense; approximately half of all German deaths from bombing occurred over the same period. It is the extravagant power of this final bombing campaign, and its massive damage to Germany’s civilian population, urban infrastructure, and cultural heritage, that has occasioned most postwar criticism of bombing strategy. To understand why the offensive was continued on such a devastating scale it is necessary to reconstruct the strategic situation as it seemed to the Western Allies in the autumn of 1944.

The most obvious answer is that there was no compelling reason after the rapid victory in France to ease the pressure on an enemy who, it was hoped, might be defeated before Christmas. There was wide popular endorsement of the bombing campaign among the home populations. The percentages in favor of bombing German civilians expressed in British opinion polls rose from less than half in 1940 to almost two-thirds by 1944, a reflection of popular anxiety to end the war quickly and a growing familiarity with bombing as a central pillar of Allied strategy.⁷⁴ The BBC air commentaries in 1944 by “Squadron Leader Strachey” (the left-wing politician John Strachey, now a temporary member of Bufton’s Bomber Directorate) had record radio audiences.⁷⁵ There was little reason for Churchill or Roosevelt to shut down the bombing offensive given the exceptional commitment to its organization and supply, and both leaders were by now eager to accelerate an end to the conflict and frustrated by an enemy whose willingness to continue fighting showed little sign of wilting. When Churchill was shown yet another political intelligence report confirming that the German people lacked the “energy, the courage or the organization” necessary to overthrow Hitler’s dictatorship, he asked to be spared any further reports on German morale.⁷⁶ From the military point of view, bombing was now part of the combined-arms offensive to defeat Germany on the ground, and although the targets were distant from the front line in eastern France, the Combined Chiefs of Staff understood that bombing would be used in general against targets that promised to expedite the army’s advance. The achievement of air superiority in the summer of 1944 required constant and vigilant defense against any prospect of German recovery, since superiority was always relative. It was feared that if the German war effort was not suppressed, the conflict might run on well into 1945, or might even reach stalemate. There was always the persistent fear, going back many years, that the German leadership might be able to turn the tide of war by jumping a stage ahead in the race for new science-based weapons.

Of all the factors that encouraged the final months of heavy bombing, the fear that the German military situation might be reversed by new weapons, secret or otherwise, kept bombers at their task. Though some of these fears might appear with hindsight as mere fantasy, the launch of the German V-weapons in the summer of 1944, and the first employment of the Me262 jet fighter/fighter-bomber, confirmed the Allied view that Germany’s military situation might abruptly improve. The so-called vengeance weapons hit British targets indiscriminately, first the V-1 flying bomb, which was first launched in mid-June, and then the V-2 rocket, which first fell on London in September. The Me262 was a crude fighter, easily adapted to mass production, but with its high speed it appeared capable of posing a serious challenge to the bomber fleets had it been available in sufficient numbers. The few units equipped with the Me262 by the end of the war claimed to have shot down 300 heavy bombers. The new aircraft was certainly welcomed by German air leaders as a possible war winner. Evaluations of the war situation produced in September 1944 and January 1945 by German Air Force intelligence presented the possibility of a “final victory” given the apparent decline in Anglo-American military capability and enthusiasm since the summer and the clutch of dangerous new technical developments available to Germany; these included the Wasserfall ground-to-air missile (V-4), proximity fuses, jet fighters (Me262 and Heinkel He162), jet bombers (Arado Ar243), rocket fighters (Me163), and equipment for nonvisual detection and destruction of enemy aircraft.⁷⁷ Some of these developments were well known to the Allies, others merely speculation. The correspondence between Spaatz and Arnold on the threat of the German jet fighters reveals the extent to which the Allies feared for the future of the bombing offensive. Spaatz wanted top priority for the development of the American jet fighter, the Lockheed P-80, and suggested the possibility that American bombers would have to change to night bombing or shallow penetration raids to keep losses to jet fighters within acceptable limits.⁷⁸ Doolittle told Arnold in August 1944 that he was not “awed” by German potential, and proposed to challenge jet and rocket-propelled fighters by head-on attack and superior turning. But by October he warned that jet aircraft and enhanced weaponry, including the powerful 30-millimeter cannon, might well “overwhelm our defenses” in attacks on Germany.⁷⁹ Anglo-American air intelligence confirmed, as Harris had warned, that by the autumn the German single-engine fighter force would be larger than at any point in 1944. Hence the arguments in favor of continued heavy bombing of German industrial and military targets.

The greatest fear was that the German leadership, like some wounded stag at bay, would unleash what are now called weapons of mass destruction. The U.S. Chemical Warfare Committee in January 1945 warned that although German leaders had not yet authorized the use of gas, the strategic situation they faced had changed for the worse: “The Germans are now fighting with their backs to the wall, on their homeland, and may out of zealousness, in defense of their own soil, or the fanatical desperation of the Nazi leadership, resort to the gas weapon.”⁸⁰ These fears had a long pedigree. Extensive preparations had been made by the RAF during the Blitz to prepare for a gas campaign against Germany. When the United States entered the war, gas warfare plans were coordinated between the two allies.⁸¹ In late 1943 these anxieties revived. Intelligence from a captured Italian diplomat suggested that Germany had large stocks of gas but would only use them “in a last resort.” Churchill, who had been the motor behind expanding Britain’s gas capability in 1941, underlined on the report the words “Gas,” “no new gas has been produced,” and “Germans would not use” as evidence of his continued concern.⁸² A report from the Analysis of Foreign Weapons Division in October 1943 to the War Department concluded that “Germany is well prepared with the necessary weapons and agents to start gas warfare at any moment,” using a variety of new toxins and means of delivery. This was indeed the case: by the beginning of 1944 the German armed forces had thousands of tons of chemical weapons on hand, including the deadly agents sarin and tabun.⁸³ In December the Joint Chiefs of Staff were supplied with full details of the toxic gases available to American forces for the conduct of gas warfare from the air in both the European and Pacific theaters, and in 1944 an accelerated program for the production of gas bombs was set in motion. In January, Arnold’s headquarters could confirm that the air force “can be effectively employed for waging gas warfare.”⁸⁴

British preparations for gas warfare were much further advanced and more likely to be used. In January 1944, Portal told Churchill that he was toying with the idea of using gas to attack preliminary V-weapon installations that had been identified in France and the Low Countries, but hesitated to do so because the repercussions of starting gas warfare “would be far-reaching.” The RAF was nevertheless alert for the first whiff of German gas in order to activate its extensive plans for airborne gas attacks. The War Cabinet was notified by the air staff the same month that if Germany should ever use it, the air force would immediately unleash six area attacks with mustard gas and two with phosgene every month. The attacks were to be divided between lighter, harassing raids and heavy concentrated raids using a mix of gas bombs, incendiaries, and high explosive, which would have to be repeated regularly “on the most densely populated centres.”⁸⁵ The air staff understood that even if the German forces used gas against the invading troops in June 1944, Churchill favored gas attacks not only on enemy troops but also on “the cities of Germany.” A list of suitable cities was drawn up in case such attacks were needed, fifteen for Bomber Command, thirty for the Eighth Air Force, and fifteen for American bombers from Italy. For the Normandy landings Bomber Command planned 11,000 sorties using gas and other bombs against a variety of military and civilian targets.⁸⁶ The stalemate in Normandy and the onset of the V-weapons campaign brought further pressure from Churchill to use gas to speed up German defeat: “I want a cold-blooded calculation made,” he wrote on July 6, “as to how it would pay us to use poison gas . . . we could drench the cities of the Ruhr.”⁸⁷ But the chiefs of staff remained opposed to the risk, while all the available intelligence suggested that there were no German plans to use it (though it was certainly discussed by Hitler and other German leaders during 1944).⁸⁸Nevertheless, by early 1945 the American Chemical Warfare Service had sufficient stocks of gas in theater to maintain a campaign equivalent to 25 percent of the total available bomblift. Almost all gas attacks would be made from the air.⁸⁹

Less well known are the plans for biological warfare against Germany developed in 1944–45. These too were the result of growing fears that a desperate enemy might utilize bacteriological warfare, possibly projected by some form of rocket propulsion. In 1942, Roosevelt authorized a War Research Board directed by George Merck, with an advisory board of prominent scientists, disguised simply as the “ABC Committee,” whose first task was to work out ways to protect the American population from a possible German or Japanese bacteriological attack. In late 1942 the work was taken over by the Chemical Warfare Service, which set up a research facility at Camp Detrick in Frederick, Maryland. As with gas warfare, intelligence began to appear in 1943 that suggested Germany was planning the use of biological warfare agents, and in particular “bacillus botulinus” (now known as Clostridium botulinum, the cause of botulism), which was impossible to detect in an airborne attack, caused symptoms in four to five hours, and death by embolism in most cases. The assumption was that Germany would only hesitate to use biological weapons because of the threat of “instant reprisal,” and as a result a program to produce lethal pathogens was accelerated. A further report in January 1944 warned that rocket or air attacks using bacteria might be imminent, and their effects “devastating.” This included the probability of using anthrax spores directed at human populations.⁹⁰ Three plants were set up for experiment and production in Mississippi, Indiana, and Utah; sixty workers were inadvertently infected, but none died.⁹¹ Before biological agents became available in sufficient quantities, which would not be before April 1945, it was recommended that retaliation should be with gas.⁹²

Would the Allies have used either gas or germ warfare? The question was never tested, since Hitler was opposed to their use and more concerned about defense against a possible Allied biological attack.⁹³ But the development of both Allied programs shows the extent to which perception of the German enemy colored the decision to continue heavy bombing in case worse weapons were to hand. The most significant factor is that fear of chemical and biological weapons prompted the Allies to think in terms of retaliation against civilian populations on a large scale, turning interwar fantasies about gas and germs into potential reality. The RAF staff thought that incendiary and high-explosive raids were more strategically efficient, in that they destroyed property and equipment and not just people, but in any of these cases—blown apart, burnt alive, or asphyxiated—deliberate damage to civilian populations was now taken for granted. This paved the way for the possibility of using atomic weapons on German targets in 1945 if the war had dragged on late into the year, a fact that is easily forgotten. Echoes can be found in the later extravagant planning for second-strike nuclear destruction during the first decades of the Cold War, when up to 80 million Soviet citizens were expected to be casualties.⁹⁴

It is against this strategic background that sense can be made of the decision to intensify the bombing offensive to be certain of securing German defeat. The summer diversion to the ground war, however, had done nothing to settle the inter-Allied arguments over bombing strategy that had surfaced in the early months of 1944. Indeed, a renewed eagerness to demonstrate what air forces were now capable of gave them fresh impetus. The first step following the decision to return the bomber forces to air force control was to change the overall command structure to more properly reflect the balance of power between the two Allied bomber forces. This was now heavily tilted toward the Eighth and Fifteenth air forces, which had on hand between them more than 5,000 heavy bombers in the European theater and could call on over 5,000 fighters, including by November around 2,000 P-51s. One major Eighth Air Force raid late in 1944 employed 2,074 bombers and 923 fighters. Against this Harris could field around 1,400 heavy bombers, mainly Lancasters, a fraction of what had been hoped for.⁹⁵ Spaatz refused to return to Portal’s direct control and preferred to retain his close relationship with Eisenhower. Under pressure from the British chiefs of staff, Arnold agreed to relinquish the link with Supreme Headquarters Allied Expeditionary Force (SHAEF), but only on condition that he would now be the representative of the Combined Chiefs of Staff for directing the American strategic and tactical air forces in Europe rather than Portal. To mollify the British, Spaatz was also formally appointed as Portal’s deputy chief of staff, together with Norman Bottomley. In practice Spaatz was left free to organize an independent campaign; he left Anderson in London and based himself in a forward headquarters next to Eisenhower, first in Paris, then Rheims. He told Arnold in late September that he preferred to keep the two forces separate.⁹⁶ Harris also understood that the change would give him once again more control over his force; since he was now the junior partner after years in which Bomber Command had been the senior manager, the guardianship of his campaign loomed larger than ever in his mind.⁹⁷

The changed command structure made a common plan for the strategic air war less likely. The Bomber Directorate in the Air Ministry in August 1944 argued for a major operation against German morale, code-named Thunderclap, for “laying on a ‘Rotterdam’” in the center of Berlin. The aim was to shatter any German hope of sustaining the war effort by “the obliteration of the visible signs of an organized Government,” using 2,000 tons of bombs dropped in accurate concentration on a 2.5-square-mile area of central Berlin. The likely shock effect on German morale was compared with the shock effect on the Dutch government in 1940, which surrendered the day after the Rotterdam raid.⁹⁸ The proposal came, for the moment, to nothing. The Combined Chiefs of Staff preferred to wait until German morale was evidently at its most fragile. The two bomber forces continued for the moment to attack what they had targeted before Normandy, Harris against city areas, Spaatz against oil and air force targets. During September, Spaatz searched for a combination of targets that would put maximum pressure on the German military war effort. Since it was now evident that the Allies, sitting on the German frontier, would not find it as easy to invade as had been hoped, Spaatz preferred a strategy that would maximize the help the air forces could give to Eisenhower. His staff worked on a program to attack major military industries and communications in the Ruhr-Rhineland, Saarland, and southwest Germany to create the maximum dislocation and demoralization of the military and administrative structures. The plan was divided in two parts: “Hurricane I” was to be directed in general at areas that contained valuable targets in western Germany during periods when visual bombing was difficult; “Hurricane II” was for precise attacks in good weather on oil installations, motor transport depots, and communications. The directive was issued on October 13 to both Harris and Spaatz, but persistent poor weather rendered “Hurricane II” unworkable, while “Hurricane I” was too amorphous in ambition and was largely ignored.⁹⁹

The search for an agreed plan did not stop the Allied bomber forces from heavy and regular attacks on German targets, almost all of them nonvisual on account of the deteriorating weather. On October 18 the two forces agreed to set up a Combined Strategic Targets Committee, chaired jointly by Bufton and Colonel Maxwell from Spaatz’s staff, to work out a better set of priorities. Air Marshal Arthur Tedder, Eisenhower’s deputy, was invited by Portal to contribute to the evaluation process. Tedder deplored what he described as a “patchwork quilt” of targets with no comprehensive pattern and recommended concentration on oil and communications as the two targets most likely to undermine German capacity to wage war. Tedder had been a firm supporter of the transport plan in Italy in 1943–44, then in the invasion of France. He was strongly supported by Solly Zuckerman, who had been responsible for working out the Italian plan, and was now advising SHAEF.¹⁰⁰ Tedder’s intervention proved decisive. Ultra intelligence in late October confirmed that attacks on transportation had already had a substantial effect on coal movements. A meeting at SHAEF headquarters on October 28 agreed on priority being given to oil targets and communications, and a new instruction, Strategic Directive No. 2, was issued on November 1 to Spaatz and Harris. The new directive, however, was a compromise, as most directives had been since the Casablanca Conference. Alongside oil and transport, the directive allowed attacks on “important industrial areas” when visual bombing was impaired, as well as policing attacks on the German Air Force organization and, when required, direct support of land operations.¹⁰¹ Bombing preferences in fact remained divided: Spaatz, Doolittle, and Bufton preferred the oil campaign; Tedder and Zuckerman sponsored communications; Harris, who had a strong personal antipathy to Zuckerman (“the ‘expert’ Mr. Solly Zuckerman”) and to Bufton (“one of my ex-Station commanders”), remained wedded to the idea that oil and transport were expensive, dangerous, and futile objectives when the destruction of cities could be more easily accomplished.¹⁰² In a famous exchange of letters with Portal (his were drafted by Bufton) in December 1944, Harris rejected the oil plan on the grounds that it would need a quarter of a million tons of bombs and months of effort to achieve it.¹⁰³

The effect of differences of opinion can be exaggerated. Allied airpower was now so overwhelming and technically sophisticated that attacks anywhere contributed to the cumulative collapse of the German war effort, and could be carried out with small losses. Harris diverted some of his area raids to the Ruhr-Rhineland synthetic oil installations in November and December, and ordered a heavy attack on Leuna-Merseburg on December 6–7 that, though deep in German territory, cost only 5 bombers from a force of 475. From 6 percent of its bombing total on oil in October, Bomber Command increased the total to 24 percent the following month. That same month the Eighth Air Force devoted 39 percent of its total on oil targets, the Fifteenth Air Force 32 percent. Despite justified fears that German efforts would be focused entirely on reconstituting oil production, the long-term trend of the oil attacks since the beginning of the year was to create a critical level of loss, whose effects on German military mobility and airpower were indeed fundamentally debilitating (see table 3.2). Most of Harris’s attacks were nevertheless devoted to night raids on urban targets, particularly on smaller cities that had not so far been the object of attack. Between October and April, when area attacks were ordered to cease, Bomber Command launched heavy attacks against cities across Germany. Some contained important rail junctions or marshaling yards, some chemical and oil plants, but most of the time the heavy bomb loads destroyed wide urban areas or continued to fall in open ground. Although a number of daylight raids were made, Harris refused to convert his force to day bombing, perhaps to ensure that the contribution of his command remained distinctive to the last.

Table 3.2: German Oil Production and Imports, 1944 (in thousands of tons)

Source: Calculated from Charles Webster and Noble Frankland, The Strategic Air Offensive Against Germany (London: HMSO, 1961), 4:516.

By contrast, American bombing, though intended to be directed at oil and transport targets, was often little distinguishable from area raiding. Much of the air policing of the German Air Force and attacks on targets of opportunity on the German transport network were carried out by fighters and fighter-bombers, which swarmed over Germany. The heavy bombers focused on major industrial and rail targets, but poor visibility for much of the winter meant that most bombs again fell widely scattered. Of every hundred bombs dropped on an oil plant, eighty-seven missed the target entirely and only two hit the buildings and equipment.¹⁰⁴ A conference on bombing accuracy called in March 1945 confirmed that most bombing since September 1944 had been blind bombing, much through complete cloud cover. From September to December 1944 only 14 percent of bombing was done with good to fair visibility, a further 10 percent visually aimed with poor visibility; 76 percent was carried out nonvisually, using a variety of electronic aids. In good visibility, at least four-fifths of bombs were found to fall within one mile of the aiming point, but through 10/10 cloud cover only 5.6 percent.¹⁰⁵ The instructions given to the Eighth Air Force in October 1944 on bombing procedure encouraged attacks in poor weather on any towns visible with the aid of H2X, for the reason that they were certain to contain some vital military targets. The effect, like the directives to Bomber Command three years before, was to encourage escalating damage to the civilian milieu and higher civilian casualties.¹⁰⁶

Whatever the operational drawbacks to flying in poor weather against heavily defended targets, enough bombs struck the oil plants and transport network to cause sufficient disruption. The transport plan was put into effect by the Eighth Air Force in early September, but serious assaults on the main rail junctions and marshaling yards began in October against Cologne, Hamm, and Duisburg. The Rhine was blocked at Cologne by a lucky strike on the Cologne-Mülheim Bridge, which collapsed into the water, blocking one of Germany’s main traffic waterways. By November the German Railway was down to eleven days’ supply of coal, by December 12 down to five days. Southern and eastern parts of Germany were starved of coal; locomotives and wagons were routinely strafed by fighters and fighter-bombers. Out of 250,000 goods wagons available, almost half were inoperable by late November. Total rail freight traffic fell by 46 percent from September 1944 to January 1945. In the Ruhr, rolling stock available for daily use was by late October half the level of September. Rolling stock was withdrawn farther away from the attacks in western Germany, but the result was to block supplies of coal and coke from the Ruhr and force a reduction of one-third in electricity generation.¹⁰⁷ Serious damage to the Mittelland Canal, the main link between the Ruhr and central Germany, left it unusable for much of October and November. Coal traffic on inland waterways was 2.2 million tons in September, but 422,000 in December.¹⁰⁸ Hitler ordered the transfer of 1,000 heavy and 2,000 light antiaircraft guns to defend key transport junctions, but as a result denuded the defense available to other vital war industries. Ultra intelligence decrypts kept the Allies regularly aware of the impact the transport plan was having and encouraged its expansion.¹⁰⁹

Somehow or other, amid the accumulating chaos of smashed rail lines, burnt-out cities, and crumpled factories, the German Air Force continued to sustain a threat to the ubiquitous enemy. Despite the long battle of attrition, there were 2,500 serviceable fighters and night fighters still available by December. Moreover, Allied losses on a number of daylight raids began to mount again: 40 bombers were lost in raids on oil targets on October 7, 40 again on November 2. But most of the raids recorded in the Eighth Air Force war diary show negligible losses and in many cases no losses to combat at all. Overall loss rates fluctuated between 1 and 2 percent throughout the period from September 1944 to the end of the war, an increasing number due to antiaircraft fire. Allied fighter losses were never high (the peak in September 1944 was only 1.9 percent of sorties); losses amounted to just 1.37 percent of all sorties in the last eight months of the war.¹¹⁰ Meanwhile, the German Air Force remained trapped in the attrition cycle set in motion earlier in the year. Combat in large formations proved dangerous even to experienced pilots. The Allied raid on November 2 that lost 40 bombers cost the German fighter force 120 planes. The collapse of aviation fuel supply played an important part; training was cut back even further and strict instructions were given on flight times and procedures to reduce fuel consumption. Both day-fighter and night-fighter squadrons found they had a surplus of pilots with available aircraft, but they could not fly because of the restrictions. The enthusiastic expectations of the Me262 jet fighter were disappointed by the slow pace of development and continued technical problems with the jet turbines. Although 564 jets were produced in 1944, the first fighter squadron armed with the new model began operations only in November.¹¹¹

The situation for the night fighters was also seriously affected by fuel shortages. Bomber Command losses fell dramatically from the high point of the summer when attacks were still suffering average losses of 6–7 percent. Over the last months of the war, loss rates dropped to an average of 1.5 percent. In 1943 a Lancaster bomber had lasted on average for twenty-two combat sorties, whereas by 1945 the figure was sixty.¹¹² The more experience crews got, the better their chances of survival. The German night-fighter force, on the other hand, was hit by the collapse in fuel supply in a number of ways. It was essential to be able to run full training programs for crews in the use of the complex scanning equipment, the SN-3 and FuG218, now available to locate the bomber aircraft. The dynamos needed to charge the radar equipment could not be operated because of fuel shortages; electricity supply to the radar stations was intermittent by the winter months of 1944, which also reduced training time on the new detection instruments.¹¹³ Most night fighters were now the high-quality Junkers Ju88G, fitted with SN2 and Naxos equipment, and the Flensburg detector used to home in to an Allied bomber’s “Monica” signals. By chance this equipment fell undamaged into British hands when a disoriented German night-fighter pilot landed in error on a Bomber Command station at Woodbridge in Suffolk in July 1944. Extensive testing soon showed that the simple expedient of turning off both the Monica tail radar and the H2S set would blind the enemy night fighters. New devices—“Perfectos,” “Piperack,” and “Serrate IV”—were developed to give warning of enemy fighters and to confuse German radar.¹¹⁴ Although a new round of research began in Germany, there was too little time or opportunity to profit from it as the infrastructure collapsed. The seesaw electronic war ended in the Allies’ favor. By 1945 the night-fighter force was a wasted asset.

In November 1944 the crisis in the German Air Force reached a peak. Göring found himself caught between two poles, Hitler’s harsh accusations over the failure of the air force and the stark reality of Allied air supremacy. He took out his own frustration by blaming his aircrew for lack of courage and loyalty. On November 11, Göring convened a tribunal (“Aeropag”) in Berlin with his senior air force commanders at which he announced that German airpower had failed and asked for solutions. It became, recorded one of those present, “a dreary forum which harped on about National Socialist influences within the Luftwaffe” but resolved nothing.¹¹⁵ The anxieties in the West about the revival of German airpower now scarcely reflected the reality. The air force relied increasingly on gestures. New Sturmjäger (storm fighter) units were created from skilled pilots who flew their aircraft, armed with heavy new 30-millimeter cannon, straight at the bomber stream, regardless of the powerful fighter escort. The suicidal tactics were occasionally accompanied by ramming, despite Hitler’s disapproval of the idea of German kamikaze. To cope with the impossibility of day-to-day combat in small formations—usually groups of ten or twenty aircraft, now directed mainly at fighter-bombers and fighter-intruders rather than the bomber stream—Adolf Galland, the general of fighters, organized a plan for a “Great Blow” by building up a reserve of fighters and fuel to release a sudden devastating attack on a large bomber stream. By November 12 there were 3,700 fighters of all kinds available, around 2,500 assigned for the blow. The object was to shoot down at least 400 bombers in one raid to try to deter the Allied offensive and buy time for the buildup of modern air equipment, “the shock the enemy needed,” one of the pilots later told his American captors, “to make them cease their inroads into the heart of Germany.”¹¹⁶

At just the point that Galland and his commanders were waiting for the weather to clear, the units were ordered westward to the Ardennes to take part in Operation Autumn Mist (Herbstnebel), better known as the Battle of the Bulge. The reserves were lost and later decimated in Operation Bodenplatte, directed against Allied airfields in early January, when almost 300 German fighters were shot down. Galland was sacked by Göring a week later on suspicion of instigating a pilots’ rebellion against his leadership. A mutiny finally surfaced in a grim confrontation at the Air Ministry after Karl Koller, the successor to Korten as chief of staff, sent a delegation to meet Göring to request changes in command, the reequipment of fighter units with the Me262 jet, and greater respect for what the fighter arm was trying to do. Göring threatened them with court-martial, but in the end the ringleaders were simply posted away from Berlin. Hitler, however, finally conceded that the jet ought to be used as a fighter rather than an ineffective fighter-bomber. Galland was sent to lead one of the first converted Me262 squadrons. On his last mission, flown on April 26, eleven days before the end of the war in Europe, he was attacked by an undetected Mustang and limped back to his base with a smashed instrument panel and both turbines damaged. As he arrived, the airfield was being bombed and strafed by Thunderbolts. He landed among them and dived into a bomb crater from his battered aircraft. Two weeks later he was explaining to his American captors the most effective way to put airfields out of commission.¹¹⁷

By January 1945 there was accumulating Ultra evidence that the choice of oil and the transport network as targets had been sensible. The stabilization of the front after the Battle of the Bulge made it evident that German forces were near the end of their fighting power, yet the assault on Germany itself promised a costly finale. The possibility of using Operation Thunderclap as a way to bring about sudden collapse was raised again. The Bomber Directorate wrote to SHAEF suggesting a bombing operation of exceptional density designed to provoke “a state of terror by air attack” in which any individual in the vicinity of the raid would realize that the chances of escaping death or serious injury “are extremely remote.”¹¹⁸ This was one of a number of voices raised over the winter in favor of punitive raids designed to spread the bombing over wider sections of the population. The Joint Intelligence Committee at SHAEF suggested in October 1944 that surplus bombing capacity before German surrender might be usefully employed in attacks against parts of Germany that had not yet been affected, “in order to bring home to the whole population the consequences of military defeat and the realities of air bombing.”¹¹⁹ Postwar interpretations of the last three months of bombing on a collapsing German war effort and a disoriented population have also come to regard the final flourish of bombing against a weakened enemy, with overwhelming force, as merely punitive, neither necessary nor, as a result, morally justified. The American air forces alone between January and April 1945 dropped more than four times the bomb tonnage used by Germany during the ten-month Blitz on Britain. For both Allied air forces the fact that it was now possible to demonstrate the full potential of airpower at a critical point of the European war played some part in their willingness to push the offensive to the maximum, in case airpower really could deliver the coup de grâce. But the calls for punitive attacks were not reflected in the prevailing directives, which still presented German resistance, particularly after the crisis in the Ardennes, as substantial enough to merit unrelenting attack. Robert Lovett wrote to Arnold early in January 1945 that despite everything, Germany showed no signs of cracking, while the German forces were fighting with such “skill and fanaticism” that it might produce “a type of dug-in, trench warfare which will be slow, costly in lives and difficult to synchronize with the increased demands of accelerated Pacific operations.” Only airpower, Lovett concluded, could break the stalemate.¹²⁰

The story of the last months of desperate German resistance is now well known, but at the time the intelligence picture for the Allies was less coherent and full of potential menace. Persistent rumors of German plans to build a “redoubt” in southern Germany or the Alps were taken more seriously than they deserved. The capacity of the Red Army to complete its victory on the Eastern Front was regarded as more imponderable than it should have been. These uncertainties help to explain the decision that led on the night of February 13–14 in the Saxon city of Dresden to a third major firestorm, which killed approximately 25,000 people in a few hours. No other raid of the war, not even Operation Gomorrah, has generated so much critical attention. Harris has regularly been blamed for conducting a needlessly destructive and strategically unnecessary raid against Dresden, but the irony is that the purpose on this occasion was dictated by the conditions of the ground war rather than the bombing campaign. It was Dresden’s misfortune to be not only in the path of the oncoming Soviet armies, but also a possible transfer route for the phantom last stand of German armies in the south. Although the city was ranked number 22 on the MEW list of target cities, with a key-point rating of 70, Harris had not yet attacked it in force, partly because of the long distance, but almost certainly because it contained no major industries linked to the current directive.¹²¹ By the autumn of 1944, Dresden was also routinely included on target lists issued to the Fifteenth Air Force stationed in Italy, along with other targets in southern and eastern Germany, but had not yet been attacked.¹²² When the Combined Strategic Targets Committee met in late November 1944, it listed cities for possible area attack when blind bombing was necessary, with an “x” to indicate oil targets present and “+” to indicate a key communications center. All thirteen cities in western Germany had one or both targets indicated; of the eleven selected in eastern Germany, seven were marked “+” but four—Dresden, Leipzig, Dessau, and Danzig—had no key target marked.¹²³

The origin of the decision to bomb Dresden has been obfuscated by the long postwar debate over who should accept responsibility or blame for what happened. The historical narrative seems, however, clear enough. The possibility of an area attack on Dresden first surfaced in October 1944 when Portal responded to a request from Churchill for a list of “area targets” that the advancing Soviet air force might be able to bomb, which included Dresden among the seven suggested.¹²⁴ Discussion about bombing cities in eastern Germany was always related to the progress of Soviet forces and the possibility of helping their advance by a display of Allied airpower. In mid-January 1945, Tedder met with Stalin to discuss the progress of the campaign against oil targets. Stalin showed great interest in the effects of bombing on German military fuel supplies and then showed Tedder the Soviet plans for the main Oder operation, launched five days later, on January 20.¹²⁵ This discussion seems to have prompted two separate responses. The JIC on January 25 announced that the Soviet offensive would be greatly helped by heavy attacks on Berlin, though priority was still to be assigned to oil targets. Portal and the air staff assessed the evidence but were unconvinced, and on January 26 their preference was still for attacks on oil and jet-fighter targets.¹²⁶ However, the same day, Churchill, who must have read the report, asked Sinclair whether there were any plans to help the Soviet offensive. Dissatisfied with Sinclair’s equivocal response, he dashed off a note on January 26 demanding to know whether Berlin “and no doubt other large cities in East Germany” were now to be considered valuable targets. Sinclair replied on January 27 that Berlin, Dresden, Leipzig, and Chemnitz were all now on the list for possible attack when the weather had improved.¹²⁷ The next day Portal wrote to Churchill that oil targets remained a key priority of the bombing war, but added the following: “We also intend, as you know, to apply as much bomber effort as we can to the cities of Eastern Germany, including Berlin: but oil must come first.”¹²⁸ Two days later Portal and Churchill both traveled to Malta for discussions with the Americans before going on to the conference at Yalta.

The second response was to set in motion actual operations. On January 27, Bottomley sent Harris the JIC report and asked him to prepare attacks on Berlin, as well as the three principal Saxon cities: Dresden, Leipzig, and Chemnitz.¹²⁹ He then drafted a paper for the chiefs of staff meeting, due to convene in Malta on January 31, which effectively summarized the grounds for the bombing:

Evacuation Areas: Evacuees from German and German-Occupied Provinces to the East of Berlin are streaming westward through Berlin itself and through Leipzig, Dresden and other cities in the East of Germany. The administrative problems involved in receiving the refugees and re-distributing them are likely to be immense. The strain on the administration and upon the communications must be considerably increased by the need for handling military reinforcements on their way to the Eastern Front. A series of heavy attacks by day and night upon these administrative and control centres is likely to create considerable delays in the deployment of troops at the Front and may well result in establishing a state of chaos. . . . It is for these reasons that instructions have been issued for heavy scale attacks to be delivered on these centres at the earliest possible moment.¹³⁰

The initiative now passed to Tedder at Eisenhower’s headquarters in Paris. After discussions with Spaatz and Bottomley, he drew up a planning document on January 31 incorporating the city attacks that would involve both British and American bombers. Spaatz, who thought that Operation Thunderclap was now the plan, preferred a heavy attack on Berlin, with high casualties, but he did not demur at a broader program.¹³¹

The only barrier to carrying out the raids was raised by the Soviet delegation at the Yalta Conference. The Soviet side demanded agreement on a formal “bombline” in eastern Germany, running through Berlin, Leipzig, and Vienna, beyond which Western air forces would not bomb for fear of hitting Soviet forces and equipment. The discussions at Yalta were resolved on February 7 by agreeing on the term “zone of limitation” to describe areas that either side could currently bomb, freeing Dresden and other cities from the Soviet proscription. It has often been argued that the Soviet side at Yalta asked for raids on Berlin, Leipzig, and Dresden, but the discussion with the Soviet chief of staff, Marshal Aleksei Antonov, recorded in the minutes, only mentions the bombing of Berlin and Leipzig; Portal seems to have insisted on including Dresden, since this was already on the list of cities suggested by the Air Ministry.¹³² Though Harris later argued at the height of the Cold War that the request to bomb Dresden had come “from the other side of the Iron Curtain,” there can be no doubt that the plan was always a Western one.¹³³ On February 7 the American military representative in Moscow, General John Deane, was notified by Spaatz that the bombing had been planned, and Soviet leaders were finally told five days later that the raid on Dresden was imminent. On February 8, SHAEF issued a formal operational instruction to Bomber Command and the Eighth Air Force to attack cities in eastern Germany when the weather was favorable.

The question for Dresden and the other cities of eastern Germany was not why they were attacked, which conformed with Allied policy on raids in support of the ground war, from Monte Cassino to Le Havre, but the way in which the raids were conducted and the weight of attack. Consistent with the new directive, Spaatz ordered a major daylight raid on Berlin on February 3, 1945, with 1,000 B-17 Flying Fortresses and almost 1,000 fighters. For once he ordered the aircraft to attack the center of the city along the lines first suggested in Operation Thunderclap, despite Doolittle’s unhappiness about the deliberate targeting of civilian areas. On the operational directive Spaatz scrawled, “Beat ’em up!” (though much later he chose to remember the raid as just another military target). The toll was high for an American raid, indeed the highest German death toll from any of the raids on the capital. An estimated 2,890 were killed and 120,000 rendered temporarily homeless. A second heavy raid on Berlin with 1,135 bombers was made on February 26.¹³⁴ On February 6, Chemnitz was also hit, by 474 American bombers; on February 14–15 a second attack was made by Bomber Command with 499 Lancasters, though cloud obscured the city and most bombs fell wide of it; a further raid was made by the Eighth Air Force on March 2. Seen from this perspective, it is evident that the raid on Dresden was made as part of a series of agreed attacks on the cities of eastern Germany. All of these raids, and not just the attack on Dresden, were undertaken in the full knowledge that these cities were filled with civilian refugees from farther east, and that their destruction was likely to cause not just dislocation but high casualties as well.

The Dresden raid on February 13–14, 1945, was carried out by Bomber Command in two successive waves with 796 Lancasters, carrying 2,646 tons of bombs (including 1,181 tons of incendiaries). Dresden’s light defenses resulted first all from the transfer of antiaircraft artillery to the Eastern Front and second from a successful diversionary raid that attracted the nearby night fighters away from the city. The first wave was not very effective, but the follow-up raid with the bulk of the Lancaster force in clear conditions achieved an exceptional level of concentration. Low humidity and dry, cold weather, combined with a very large number of small fires quickly started, proved ideal conditions for the generation of another firestorm. The flames consumed fifteen square miles of the city, an area that exceeded the damage at Hamburg. Recent estimates from a historical commission in Dresden have confirmed that the original figure suggested by the city’s police president in March 1945 of approximately 25,000 dead is the best available estimate. Out of 220,000 homes, 75,000 were destroyed.¹³⁵ The firestorm, like the Hamburg conflagration, left bodies mummified or reduced to ash, making the final count difficult. A further 1,858 skeletons were unearthed when the city was slowly rebuilt after 1945. The aiming point in this, as in all area attacks, was the historic city center, which was entirely burnt out. The next day the Eighth Air Force carried out its first raid on the marshaling yards of the city, but the smoke from the previous night’s bombing obscured the target and the 700 tons of bombs destroyed more of Dresden’s hapless streets. In the afternoon 210 B-17s, unable to bomb their primary oil target, blind-bombed the city with another 461 tons. In all, almost 4,000 tons of bombs were dropped on a single target in less than twenty-four hours.

Unlike any of the other major raids in the last months of the war, the Dresden attack had immediate repercussions on Allied opinion. Two days after the raid an RAF officer at SHAEF headquarters gave a news conference in which he talked about bombing cities deliberately to cause panic and destroy morale. An Associated Press correspondent, Howard Cowan, filed a report successfully past the SHAEF censor, and by February 18 the American press was full of the news that the Allies had at last decided “to adopt deliberate terror bombing.” Arnold was compelled to run a campaign to reassure the American public that Dresden had been attacked, like Chemnitz, as a major communications center, entirely consistent with American bombing policy.¹³⁶ It was hard to stifle the debate. Goebbels released to the neutral press news that 250,000 people had been killed in Dresden (by the judicious addition of an additional zero to the provisional casualty estimate). In Britain and America news of the death toll was soon public knowledge. The Bombing Restrictions Committee in London publicized the figure of 250,000 at once and provoked a furious correspondence accusing the committee of acting as the mouthpiece for German propaganda. Air Ministry statements in the House of Commons dismissed the accusations of terror bombing by claiming that no one, air marshals or pilots, was trying to work out “how many women and children they can kill.”¹³⁷But for the first time the real nature of area and blind-bombing attacks came under public scrutiny.

This may explain Churchill’s now well-known decision to send a minute to Portal on March 28, 1945, protesting that the policy of bombing “for the sake of increasing the terror, though under other pretexts, should be reviewed.” He asked Portal to focus on oil and transport, as the strategic directives had intended, instead of “mere acts of terror and wanton destruction.” Harris was told of the document by Bottomley, who suggested that Churchill might have been worried about the shortage of German building materials, but Harris was outraged. He replied that city bombing had always been strategically justified because it would shorten the war and save the lives of Allied soldiers, an assertion difficult to reconcile with the five long years of British bombing. Portal persuaded Churchill to moderate his original minute for the chiefs of staff, which he did, but area bombing’s days were now numbered. Churchill did not indicate his motives, and the entire episode of Dresden is missing from his history of the Second World War. It is possible that the publicity surrounding bombing as a result of Dresden worried Churchill as he contemplated a general election at some point in the next few months; it probably reflected his persistent ambivalence about bombing ever since its first disappointments in 1940 and 1941; or it may be that he finally realized, as Allied forces now poured into the broken cities of the Ruhr, just what bombing had done (on March 26 he lunched on the banks of the Rhine with General Bernard Montgomery, commander of the British 21st Army Group) and was affected by its enormity, as he had been when he wandered through British cities during the Blitz. Harris much later in life dismissed the episode as unimportant; he told his biographer that Churchill’s attitude to him did not alter “in any perceivable way” between 1942 and 1945. But the rift was important enough to be suppressed until its publication in the official history in 1961.¹³⁸

Whatever Churchill’s misgivings, British city bombing continued in ways that were evidently punitive in nature and excessive in scale. Just ten days after Dresden, Bomber Command attacked the small town of Pforzheim. The marking worked well and the bombers dropped their loads from just 8,000 feet (instead of 18,000–20,000 feet on raids against defended targets); the subsequent conflagration consumed 83 percent of the city area, until then the worst in any raid of the war, and killed an estimated 17,600 people, though the death toll, the third highest in the European bombing war, has never had the publicity accorded to Dresden. The ruins of Cologne, hit by more than 250 wartime raids, were raked over again by a massive Bomber Command attack on March 2 by over 700 Lancasters, just four days before it was occupied by American forces. Essen suffered the same fate on March 11, with a macabre finale by over 1,000 bombers dropping 4,661 tons on a desolate landscape only hours before it fell to the advancing army. On March 24, Bomber Command headquarters portentously announced that, thanks to bombing, the “Battle of the Ruhr,” then in its last furious days of ground combat, “is already over—and Germany has lost it.”¹³⁹ On March 16–17, 1,127 tons of bombs were dropped on the small medieval city of Würzburg, killing between 4,000 and 5,000 people and destroying 89 percent of the city, a wartime record. Hildesheim was half destroyed on March 22 (the town center “should make a good fire,” the crews were told).¹⁴⁰ The small city of Paderborn was destroyed on March 27, and half of Plauen on April 10–11. The final catalog of area attacks could not be restrained even by Churchill. On April 4, Portal, spurred perhaps by Churchill’s minute, had notified the chiefs of staff that area attacks on industrial districts for the sake of destruction would now cease. When Harris destroyed Potsdam in a devastating raid on April 14–15, Churchill wrote angrily to Sinclair, “What was the point of going and blowing down Potsdam?” Portal replied a day later assuring the prime minister that Harris had already been told to discontinue industrial area attacks.¹⁴¹ The directive sent to Harris on April 16 for the first time since February 1942 no longer contained industrial areas or morale as dedicated objectives.¹⁴²

The American air forces wound down operations in April 1945. Much of the bombing since the February attacks was tactical in nature, directed at almost any target that could be deemed an element of German resistance. Operation Clarion was carried out with mixed success against a range of smaller communications targets. On April 5 all objectives were defined henceforth as tactical, but American bombing of the shrinking German area reached a crescendo, with 46,628 tons dropped in nineteen days of raiding, almost the same weight dropped during the German Blitz, but in just three weeks. The last raid by the Eighth Air Force was made on April 25 against the Skoda works at Pilsen; the last by the Fifteenth was on April 26 against the Austrian city of Klagenfurt.¹⁴³ Spaatz attended the surrender ceremony in Berlin on May 8 as the senior air commander in Europe. The Soviet delegation, however, refused to allow him to sign as the equal of Marshal Zhukov, the conqueror of Berlin, and he had to add his name underneath as a witness.¹⁴⁴

Spaatz already knew that the Eighth Air Force was destined to go to the Pacific under his command to help complete the defeat of Japan. British bombers were also expected to contribute, and preparations were in hand to undertake operations against Japanese cities that had already been reduced to ash in a series of extensive incendiary attacks carried out by the former Eighth Air Force divisional commander, Curtis LeMay. RE8 produced a report on May 25, 1945, two weeks after the German surrender, titled “Area Attack Against Japan,” recommending that since everything easily combustible had already been burned down, Bomber Command should use 4,000-pound blast bombs to destroy any urban areas or industrial targets still standing. From previous analysis carried out on the vulnerability of Japanese housing, it was calculated that each bomb would destroy more than ten built-up acres, whereas in Germany the figure had been only 1.5.¹⁴⁵ The air war in Europe was over, but Japan was soon to profit from its grim lessons.

Surveying the Wreckage, 1945

In August 1944, Spaatz had asked his air force commanders to speed up the defeat and surrender of Germany so that a special committee could review what bombing had achieved in Europe in order to apply their conclusions to the war against Japan.¹⁴⁶ The idea of undertaking a serious scientific survey of the bombing campaign had first been aired in the spring of 1944 and was enthusiastically supported by Spaatz, who approached Arnold and Lovett on the subject in April. Arnold wanted an independent assessment of the question “Was strategic bombing as good as we thought it was?” With Lovett’s strong support, the air force put together a plan that they presented to the president in September. Roosevelt approved the project and asked the secretary of war, Henry Stimson, to establish the new office. Arnold chose a businessman, Franklin D’Olier, president of the Prudential Life Insurance Company, to head a board of professional economists, academics, and analysts, and on November 3, Stimson formally set up the United States Strategic Bombing Survey (USSBS), based in London; a forward base was set up after the end of the war in the resort town of Bad Nauheim, with branches in other German cities. Approval was given to enlist 300 civilians and 850 officers and men from the armed forces.¹⁴⁷ Their task was to produce comprehensive reports not only on the results of American bombing but also on the RAF offensive. The survey began its operations before the end of the war, as German territory was gradually captured.

The RAF also began to plan for a possible survey in the spring of 1944. The British side assumed that they would collaborate with the Americans, and on August 10, 1944, the chiefs of staff authorized the Air Ministry to prepare an inter-Allied survey organization. Arnold was solidly opposed to any joint venture, though it did not stop the USSBS from commenting at length on British bombing. By the time Sinclair finally proposed a survey to Churchill in December 1944, it was to be a British project. Churchill brusquely dismissed the idea of what was now called the British Bombing Research Mission, partly because of the assumption that it would take at least eighteen months to report, and hence be of no use in the war against Japan, but also because he deprecated tying up “the use of manpower and brainpower on this scale.”¹⁴⁸ Instead of the large staff envisaged by the ministry, Churchill recommended a limited group of twenty to thirty people. His intervention invited months of bureaucratic wrangling over who should take part and at what cost, until Portal finally lost patience, abandoned the idea of the mission, and recruited a small unit already established at SHAEF, the Bombing Analysis Unit, as the core of a British Bombing Survey Unit (BBSU). The new organization was formally launched on June 13, 1945, months after the American survey had begun its work.¹⁴⁹ The unit was to be run by an air force officer, Air Commodore Claude Pelly, and the SHAEF target adviser, Solly Zuckerman, assisted by staff from the RE8 division of the Ministry of Home Security, which was to be closed down when the war ended.¹⁵⁰ Both men were committed enthusiasts of the attack on communications, and their work and the subsequent reports reflected their bias. Their terms of reference were to examine the effects of bombing on German fighting capacity, the effectiveness of German defenses, and the accuracy of assessments of damage.¹⁵¹ Already hostage to the small size and limited resources of the new unit, the BBSU became a vehicle for Zuckerman to argue his transport case in contrast to the disinterested analysis sought by the American air force. Much of the work of the BBSU was reliant on American research and expertise, a reflection of the rapid shift in the balance of power between the two air forces.

The process of collecting files and statistics and interrogating senior German personnel began at once. By the end of May a great many of the key figures had already been interrogated, including Göring, whose transcripts reveal an almost boyish desire to share his knowledge of the German Air Force with the victors. The provisional conclusions among the cohort of German airmen, engineers, and ministerial staff subject to interrogation were almost unanimous. A British intelligence assessment, “Factors in Germany’s Defeat,” produced by May 17, included an interrogation with Adolf Galland, who ranked the offensive against transport, then oil, then the air force as the most decisive.¹⁵² In mid-June a full report of interrogation extracts was produced by the director of American air force intelligence at SHAEF, George C. McDonald. They also showed that the three critical targets were considered to be oil facilities (“The general opinion of the German leaders is that the attack on synthetic oil was the decisive factor”), communications (“brought about the final disruption of the German war effort”), and the German Air Force—achieved through attacks on aircraft production, airfields, and combat attrition.¹⁵³ Göring thought the collapse of oil supply to be the single most critical factor—“without fuel, nobody can conduct a war”—while Albert Speer, Hitler’s minister for armaments and war production, ranked communications at the top of the list of critical targets. Erhard Milch, Göring’s deputy at the Air Ministry, ranked “synthetic oil plants and railway communications” together.¹⁵⁴ On area bombing the German judgment was largely negative. It did not “cause the collapse of the German people” and was regarded, according to McDonald, as “the least important of the major target complexes.” When Göring was asked in one of his first interrogations on May 10 whether precision or area bombing was more effective in Germany’s defeat, he replied, “The precision bombing, because it was decisive. Destroyed cities could be evacuated but destroyed industry was difficult to replace.”¹⁵⁵ In a USSBS interview on May 24 with Karl Koller, the last German Air Force chief of staff, Koller claimed, not altogether plausibly, that without precision attacks “Germany would have won the war.” He confirmed that oil and transport facilities were fatal targets for Germany.¹⁵⁶

In general, Allied assessments reached the same conclusion. The USSBS produced over 200 detailed reports on every aspect of the bombing war, but the “Over-all Report” reflected the views of those interrogated. The survey board had an interest in arguing that in the Western theater airpower was decisive, thanks chiefly to the air victory achieved over Germany in the spring and summer of 1944, “which made devastating attack on [the German] economy possible.” The report highlighted the relative failure of area attacks, which “had little effect on production,” while singling out oil and communications as critical. The attack on Ruhr steel in late 1944 was also added as a key factor, but the choice of this period rather than Harris’s “Battle of the Ruhr” in 1943 added weight to the implication that it was American bombing that had been decisive. The treatment of city attacks (4 pages out of 109) minimized their impact on economic output. Statistics were presented showing that city attacks, overwhelmingly by the RAF, cost only around 2.7 percent of German economic potential in the target areas. It was calculated that the combined offensive cost 2.5 percent of potential German output in 1942, 9 percent in 1943, and 17 percent in 1944 (figures that were roughly consistent with the claim that 5 percent of British output was lost during the lighter Blitz). Since area bombing in 1944 experienced diminishing returns, ton for ton, by dropping on previously destroyed areas, the implication again was that most of the production loss was due to American bombing of selected target systems.¹⁵⁷

Perhaps more surprisingly, this was the conclusion also arrived at by the BBSU when its main report was finally completed in draft in June 1946, almost a year after that of the USSBS. There were months of delay in arguments with the senior commanders who had been responsible for the offensive, except Harris, whose views were not canvassed. Unlike the USSBS reports, which were easily available, the BBSU final survey and subsidiary reports were given only a limited circulation. The final report was critical of almost all phases of Bomber Command’s activities except the final phase against oil and communications targets. A good deal of the report was devoted to demonstrating that the final industrial and military crisis in Germany was a result of the disintegration of rail and water traffic: “Enough has been said to show that the collapse of the German transport system . . . was the fundamental and main reason for the contemporaneous collapse of German war industry.”¹⁵⁸ This fitted with Zuckerman’s own prejudices. Even the assessment of oil supply, which the report regarded as critically disabling, concluded that the offensive against transport was responsible for preventing the recovery of Germany’s oil position.¹⁵⁹ The judgment of the report on area bombing of German cities was even more damning than that of the USSBS. Using methods pioneered earlier by the RE8 department, Zuckerman’s team calculated on the basis of twenty-one heavily bombed industrial cities that area bombing reduced potential war production by 0.5 percent in 1942, 3.2 percent in the first six months of 1943, 6.9 percent in the second six months of 1943, and then approximately 1 percent throughout 1944, when bombing again brought diminishing returns and area bombing was only one of the factors affecting output. The figures were lower than the USSBS estimates (which had been speculative extrapolations) because they were based on careful research across a range of cities and because they measured potential loss against a rising trend of output. In all twenty-one cities studied, war production expanded faster than it had done in a control cohort of fourteen cities not subject to attack.¹⁶⁰

Damning though this indictment was, and partial though Zuckerman’s position appeared to be, it fitted not only with the interrogation evidence but with the views among Air Ministry officials and RAF commanders in the two years after the war when hard thinking had to be done about what had been achieved by Bomber Command, rather than the combined offensive as a whole. Sydney Bufton, once an advocate of incendiary attacks on cities, produced a long critical assessment of area bombing in January 1945, in which he admitted the failure at Hamburg in 1943 as an example of misplaced confidence in the economic or morale effects of heavy urban destruction.¹⁶¹ Norman Bottomley, Portal’s deputy for the last three years of the war, and Harris’s successor as commander in chief at Bomber Command, contributed an assessment of British bombing at a workshop organized by Tedder, now chief of staff, in August 1947 under the code name Exercise Thunderbolt. The effect of area attack, he concluded, was “great but never critical,” nor was enemy morale ever “critically undermined,” a fact he blamed on poor intelligence. “Offensive against oil and transportation proved most effective,” he wrote, but only after the achievement of the vital precondition of air superiority.¹⁶² A lecture given in 1946 by one of the RAF officers on the British survey highlighted air force attrition, oil, and transport again, but argued that “little worthwhile” had been achieved by area attacks before 1943, and thereafter the resistance of the German population and the reserve capacity of German industry made them “resilient to area attack.”¹⁶³ Given the uniformity of opinion on both the German and Allied side, the one based on experience, the other on extensive research, it is surprising that the effects of bombing have occasioned so much debate ever since. The proximate causes—defeating the German Air Force and emasculating oil supply and transport—are unlikely to be undermined by further research.

The statistics nevertheless require some explanation about why the overall impact of bombing for much of the war period should have been so much lower than expectations. At its simplest level, as Henry Tizard put it after the war, “You can’t destroy an economy.”¹⁶⁴ American economists drafted in to advise the U.S. war effort in Europe were critical of the idea that bombing either areas or specific industries would of itself produce cumulative damage. The Hungarian émigré economist Nicholas Kaldor, a member of the USSBS team, argued that the critical factors in choosing economic targets were the degree of “cushion,” the degree of “depth,” and the degree of “vulnerability.” The first was governed by the existing elasticity of the economy in terms of finding additional or substitute resources for those lost to bombing; the second measured the extent to which a particular product or resource was close to actual military use, since the farther back in the production chain, the longer the time before bombing would affect military performance; the third was governed by the extent to which concentrated, and relatively inflexible, capital industries could be effectively destroyed from the air.¹⁶⁵ Kaldor and his economist colleagues argued that for most of the war period Germany had a large cushion of resources of capital stock, labor, and raw materials that could be allocated to sustaining war production. His conclusion was based partly on the assumption, now generally regarded by historians as invalid, that Hitler did not order full-scale mobilization until 1944. The degree of allocation of productive resources to war purposes was in fact high from the start of the war, but many of the economies of scale characteristic of large-scale industrial production became effective only by 1942–43, while the unanticipated length of the campaign against the Soviet Union distorted war production plans at a critical juncture in 1941–42.¹⁶⁶

Yet Kaldor was not wrong to argue that a cushion existed. The index of armaments output showed that German production increased threefold between 1941 and 1944, despite all the bombing; some individual categories of weapon expanded more than this, fighter aircraft by a factor of thirteen, tanks by a factor of five, heavy guns by a factor of four.¹⁶⁷ As a result of the conquest of much of continental Europe, Germany had access to large resources beyond her borders. Although this also involved economic costs to Germany, occupation meant that over 119 billion marks were contributed to Germany’s war budget, one-quarter of all the costs of the armed forces; 7.9 million forced workers and prisoners of war were compelled to work in Germany, while an estimated 20 million more worked on orders for the German war effort in the occupied zones.¹⁶⁸Moreover, German technical and organizational ingenuity made it possible to find substitute products or productive capacity even for “bottleneck” industries like ball bearings, where, as Kaldor argued, the target had “run away” by the time the Allies attacked it again in 1944.¹⁶⁹ The German economy, wrote the USSBS economist John Kenneth Galbraith in an early evaluation, was “expanding and resilient, not static and brittle.”¹⁷⁰

For most of the Allied bombing offensive these factors were either insufficiently known or not understood, and bombing, as a result, was relatively ineffective. Only in 1944, with the American decision to focus on enemy airpower, oil, and transport, were three targets chosen that fortuitously matched Kaldor’s calculation. The attack on the German aircraft assembly industry, as part of the assault against enemy airpower, was the least successful because of the substantial cushion that existed in dispersing the final stages of production; all German leaders claimed, however, that repeated attacks on aero-engine production would have been critical. Oil and transport facilities, on the other hand, had poor cushioning possibilities once heavy attacks began, were highly vulnerable to sustained attack, and had a positive “depth” factor because both were needed almost immediately by the armed forces and by industry to sustain fighting capability and output. The campaign against the German Air Force was indeed a precondition for the success of the campaigns against oil and transport, and was a direct result of the changing tactics of day bombing and the high priority given by Spaatz to suppressing German airpower as fully as possible. When German oil installations and air force operations both threatened a limited revival in late 1944, Spaatz shifted once again to priority oil and counterforce attacks. It is difficult not to argue that the U.S. air forces had a surer strategic grasp and a clearer set of strategic objectives than did Bomber Command. Counterforce operations and the search for target systems that would unhinge the enemy’s military efforts were central elements in American wartime air doctrine. The RAF, by contrast, thought of airpower more as a form of blockade, and was never enthusiastic about counterforce operations or attacks on transport, though both had been adopted in the Mediterranean campaign. The defeat of the German Air Force over Germany and the massive dislocation of German transport were primarily American achievements.

Area bombing was nevertheless, despite its critics, not entirely without impact on the German war effort. The random and scattered nature of much of the city-bombing campaign had evident opportunity costs for the German war effort, in addition to the effects of substantial civilian casualties and damage to housing. Consumer goods production had to be increased in 1943, against the trend of total war mobilization, to meet the needs of bombed-out families. The night bombing interrupted utility services and hit shops and occasionally factories, necessitating the allocation of additional resources of manpower to cope. There is no way in which these kinds of costs can be computed, any more than there was in Britain as a result of the Blitz. The real question concerns assessment of the damage done to the German industrial working class, since this was the whole rationale behind the campaign. It has never proved possible to calculate the number of workers killed, rather than nonworkers (elderly, women with families, children, etc.), but some sense of the limitations of any such assessment can be found in the death statistics in Hamburg, where on the night of the firestorm in 1943 that killed over 18,000 people, only 280 were killed in the factory district, away from the main area of bombing.¹⁷¹ Workers were not always the most likely victims, but even if the estimated total of 350,000 German dead from bombing were all workers, that would still have represented only 1.6 percent of the German industrial and rural workforce, some of whom would have been killed by American daytime bombs rather than by the RAF.

The other argument for area bombing was the high level of absenteeism it would induce, though the British evidence, on which the strategy was based, was scarcely convincing. German records show that absenteeism as a direct result of bombing made up 4.5 percent of hours lost at the height of the bombing in 1944; an additional 10.8 percent of hours were lost due to illness or leave, though these may well have been a response to circumstances caused by bombing. Figures of hours lost due to bombing were higher in targeted industries (7.9 percent in shipbuilding, 10.6 percent in vehicle production), but much of this loss was the result of precise attacks by day rather than by Bomber Command at night.¹⁷² The state of “morale” among the German population, which was also a stated objective, will be examined in greater detail in the next chapter.

The emphasis that has usually been put on the economic impact of bombing, in part a result of the very full economic data supplied by the bombing surveys, has had the effect of avoiding the more important question about the effect of bombing on the German military effort. Here the impact is more evident, though it would be prudent not to take at face value the USSBS claim that bombing was decisive. The air war over Germany was, in Albert Speer’s phrase, “the greatest lost battle.”¹⁷³ But it was a battle that was won alongside the Allied armies and navies; no particular service was decisive on its own. At the end of the war, Bufton observed that the whole purpose of the bombing offensive in 1944 “has been designed to weaken the German war machine as a whole so that it could not resist successfully when the Allied Armies made their final attack.”¹⁷⁴ One of the principal criticisms of the BBSU report made by Portal (now Lord Portal of Hungerford) was the failure of its authors to grasp that “from 1941 onwards, if not before, the object of the bomber offensive from the U.K. was to weaken Germany to such an extent that an invasion of the Continent would succeed.”¹⁷⁵ The problem with such claims is to be able to find a way to calculate the extent to which bombing really did inhibit German fighting power. As in the British case, the critical factor was the distorting effect that bombing had on German strategy once it became necessary to divert large resources to the military combat against the air offensive. Bombing, as Speer recognized, really did come to constitute a “Second Front” by 1943, preventing German military leaders from using airpower effectively at the fighting front as they had done in all the campaigns from 1939 to 1941. Failure in Russia, in the Mediterranean theater, and against the Allied invasion of France owed a great deal to the fact that German fighter aircraft, guns, ammunition, and radar equipment were tied up in the Reich. This had not necessarily been the Allied intention, which focused on unhinging the domestic war effort, but it undoubtedly contributed to the military outcome at the fighting fronts throughout the last three years of war and compensated for whatever weaknesses might have existed in Allied combat experience or skills.

There are two ways in which the effects of bombing on Germany’s military effort can be directly measured. The combined offensive distorted German military strategy by imposing a heavy cost in active and passive antiaircraft defense. One of the keys to Germany’s early battlefield successes was the employment of fighters, fighter-bombers, and medium bombers in support of ground forces. The Allied bombing forced the German leadership to switch aircraft back to the defense of the Reich and to reduce sharply the proportion of output devoted to frontline bombers and fighter-bombers, as table 3.3 demonstrates. This had the immediate effect of limiting severely the offensive airpower available on the battlefield.

Table 3.3: German Fighter and Bomber Strength and Production, 1943–44

*Figures for January 1944.

Source: Calculated from Charles Webster and Noble Frankland, The Strategic Air Offensive Against Germany (London: HMSO, 1961), 4:494–95, 501–2.

In early 1943, 59 percent of German fighters were in the Western theater facing the bombing; in January 1944, 68 percent; by October 1944, 81 percent. At the beginning of 1944 German aircraft available on the Soviet front were little more than the number a year before, in the Mediterranean theater they were 40 percent fewer, but in defense of Germany the number increased by 82 percent. The same was true for the distribution of antiaircraft guns: in the summer of 1944 there were 2,172 batteries of light and heavy antiaircraft artillery on the home front, but only 443 batteries in the Mediterranean theater, and 301 on the whole of the Eastern Front.¹⁷⁶ This situation left German armies denuded of air protection at a critical juncture of the ground war on the Eastern and Mediterranean fronts, while the diversion to the defense of Germany created just the conditions the American air forces needed to be able to overcome the German Air Force in the “Battle of Germany,” perhaps the single most significant military achievement of the offensive.

For the German war effort the costs of all forms of air defense by 1943–44 were substantial in terms of both manpower and equipment. The antiaircraft service absorbed 255,000 people in 1940, but 889,000 at its peak in 1944; the 14,400 heavy and 42,000 light guns by 1944 required production of 4,000 new guns a month, and antiaircraft units consumed one-fifth of all ammunition, half the production of the electronics industry, and one-third of all optical equipment.¹⁷⁷ The passive civil defense personnel numbered around 900,000 (supported by 15 million members of the Air Defense League); the numbers involved in post-raid clearance fluctuated over time, but they totaled by 1944 in the hundreds of thousands. Civil defense and medical equipment had to be maintained in the face of wide losses, hospitals had to be built and repaired, and fire services expanded. Few of those involved would have been potential soldiers, but many would have been potential war workers, if they were not already combining civil defense activities with paid work. This does not mean that civilians were ipso facto legitimate targets, but as in the British case, it shows that bombing compelled German resources to be allocated in ways that directly affected German military potential at the fighting front and the pattern of strategic choices. Without bombing, Germany would have been as free to optimize the use of resources and to conduct the military war effort as was the United States. The military consequences of the bombing campaign were clearly more important than the economic, psychological, or political ones.

This still begs the question of what it meant for the Allies. Galbraith, one of the USSBS team, later wrote in his memoirs that the man-hours, aircraft, and bombs “had cost the American economy far more in output than they had cost Germany.”¹⁷⁸ Both air forces were sensitive to this charge and calculated themselves what proportion of the national effort could be attributed to the bombing. In the British case the proportion was calculated to be 7 percent of all man-hour equivalents, in the American case an estimated 12 percent of wartime expenditure, both figures that did not distort exceptionally the structure of either war effort, unless account is taken of just how wasteful much of the bombing was.¹⁷⁹ Moreover, these were positive choices made about the allocation of resources, where in the German case the choice was involuntary, an addition to the other choices made about the distribution of strategic resources. Nevertheless the cost in manpower and aircraft lost in combat was substantial. In RAF Bomber Command, 47,268 were killed in action (or died as prisoners of war) and 8,195 in accidents. According to Harris, an estimated 135,000 flew in combat with Bomber Command, a loss rate of 41 percent. Total RAF dead during the war from all causes totaled 101,223, so that Bomber Command deaths amounted to 54.7 percent of all RAF losses.¹⁸⁰ Of these, the largest non-British contingent was composed of Canadians, 9,919 of whom died in Bomber Command.¹⁸¹ Total wastage of Bomber Command aircraft from all causes was 16,454.¹⁸² American heavy bomber losses against Germany totaled 10,152 between 1942 and 1945, and the total killed in all theaters against Germany was 30,099.¹⁸³

Balancing the Allied losses against the German figures for aircraft and personnel says little about the final outcome. The costs were modest compared with the 9 million Soviet military dead and 5 million German dead, reflecting the priority of both Western Allies to avoid repeating the losses of the Great War for publics likely to be less tolerant of the escalating human cost to themselves than were populations under dictatorship. For Britain and the United States, the political advantages of preferring bombing to other forms of combat were to be found in the desire to limit the cost to the home population while maximizing the use of advanced technology and large manufacturing capacity to impose insupportable costs on the population, economy, and military structure of the enemy. Bombing could be used to maintain domestic morale and to exert leverage on the enemy in ways that were rendered easily visible in the democratic media. That the campaign could have been conducted differently, at lower cost (to both sides), and with greater efficacy is not in doubt, but it is evident from the historical record why these opportunities, strategic and technical, were missed or ignored or misunderstood, or incompetently attempted. War is always easier to fight looking backward.