Chapter 6

Bombing Friends, Bombing Enemies: Germany’s New Order

In early 1944 the U.S. Eighth Air Force published a widely circulated publicity booklet, Target: Germany. It purported to tell the story of the first year of the American bombing of the German enemy, “raining havoc and destruction on the Nazi war machine.” The inside covers show a map of Europe where the force’s bombs had fallen: there are nineteen German targets but forty-five in France, Belgium, and the Netherlands. For much of the first year the apprentice American force took the short route across the Channel to bombard military-economic installations working for the German war effort. Most of the photographs in the richly illustrated text are of raids made on France and the Low Countries. The first raid on German territory finally took place late in January 1943, but more accessible European objectives were still seen as a useful way to get the crews to cut their teeth on combat.¹

The bombing of European targets outside Germany and Italy was in reality more complex than this and was large in scale. The occupied territories of western and northern Europe—France, Belgium, the Netherlands, Norway, and Denmark—absorbed almost 30 percent of the bomb tonnage dropped by the American and British bomber forces. The occupied or satellite countries in eastern Europe and the Balkans absorbed another 6.7 percent.² Well over one-third of all Allied bombs dropped on Europe fell on the German New Order, making the experience of bombing in the Second World War a European-wide one. The purpose behind these bombings, and their consequences for the populations caught in the coils of German expansion, are seldom treated as systematically as accounts of the bombing of Germany, yet they cost at least 70,000–75,000 lives, most of them among peoples sympathetic to the Allied cause. The majority of those losses, and most of the bomb tonnage that caused them, occurred in western Europe, principally France. These areas were near enough to reach and in 1944 provided the territorial springboard for the Allied invasion of the western half of the German empire. Much of the bombing in the later period was in a loose sense tactical, intended to achieve a direct military end for the ground forces; but much of it was long-distance and heavy, designed to nibble away at war production for Germany in occupied territory, but also to promote wider political aims. According to the British Political Warfare Executive, set up in 1941, bombing of occupied areas promoted both “morale breaking” and “morale making.” Collaborators and Germans would be demoralized by the experience; those who did not collaborate would be encouraged at the prospect of liberation.³ To be bombed in order to be free now seems paradoxical, but the policy governed much of the bombing that spread out across the entire European continent between 1940 and 1945.

Disordering the “New Order”

The rapid German victories between 1939 and spring 1941 brought most of continental Europe under German control. Neutral states were compelled to work with the changed balance of power, while those states that were allied with Germany—Slovakia, Hungary, Romania, Croatia, Bulgaria—were satellites of the powerful German core. In Berlin the sudden transformation in German fortunes brought a flood of plans and projects for a European New Order that would secure Germany’s permanent political and economic hegemony. The conquered states became involuntary participants in this larger project, compelled to provide economic resources, finance, and labor for the German war effort, and doomed to be the battlefields on which the enlarged German empire would defend its borders. Right to the end of the war there were still German forces fighting in the Netherlands, Italy, Hungary, and the Czech lands, alongside the defense of the German homeland.

It was inevitable under these circumstances that Britain, and later the United States, would have to engage the enemy on the territory of occupied or satellite states where German forces were dispersed. Until the Western states were in a position to mount a major land invasion, airpower was regarded as the principal means available to attack Axis military resources across Europe and to undermine the extended war economy established throughout the German New Order. Because of the problem of aircraft range and the danger of flying for long periods over heavily defended territory, it was only possible in the first years of war to attack targets in western and northern Europe. Eastern and southeastern Europe were only struck regularly from late 1943 onward, chiefly by the American Fifteenth Air Force based in southern Italy. An alternative was to rely on local resistance and sabotage, and throughout the period in which air forces bombed targets in New Order Europe, the Allies tried to encourage the occupied peoples to take a hand in their own liberation even while subject to regular bomb attacks. Throughout the war period this resulted in awkward decisions for the Allies about the level of damage that should be inflicted on targets situated among potentially friendly civilian communities forced to work for the German war machine or, in some cases, voluntarily collaborating with it. The erosion of ethical restraints on bombing German industrial cities was a simpler issue than the moral dilemma of causing civilian casualties among those held hostage by German military success.

The debate about bombing friends as well as enemies began as soon as Britain was expelled from the Continent and France defeated. In July 1940 the War Cabinet agreed that any military target could be bombed in the northern and western parts of France occupied by German forces (though not in the unoccupied zone ruled from the new government seat at Vichy).⁴ The problem was to decide what counted as a military target since it was already assumed that in the German case this meant industry, utilities, and worker morale alongside more evidently military objectives. On August 17, Air Intelligence provided a list of what were defined as “Fringe Targets” around the edge of occupied Europe that could be subjected to air attack. The fringe included targets up to thirty miles inland in Norway, Denmark, the Netherlands, Belgium, and France. In Scandinavia there were twenty-five targets, consisting chiefly of oil installations and airbases (but including the Norwegian port of Kristiansund, already heavily bombed by German aircraft); in the Low Countries sixty-one targets were identified, ranging from electricity-generating stations to iron and steelworks; in France thirty-one targets were listed around the coast from Dunkirk to Bordeaux, including an aero-engine works at Le Havre, a power station at Nantes, and a marshaling yard at Lille, a little over thirty miles from the coast.⁵ Over the course of the autumn, additional target information was processed and detailed target maps supplied. The list for France expanded to fifty-eight objectives located in the thirty-mile zone: nine oil installations, eight chemical plants, eleven aircraft works, seven blast furnaces/steel mills, eleven shipbuilding firms, and another dozen smaller targets. The targets were given star ratings to indicate their importance, three stars for the highest priority, of which there were seventy-seven by the spring of 1941.⁶ In May 1941 it was agreed that the RAF could undertake attacks on “deep penetration targets” where these could be reached easily by day without excessive risk.⁷ Conscientious anxieties played little part in these early raids. Escalation was soon built into the process of deciding what could be hit and under what conditions.

Many of the early raids around the fringe were tactical bombing operations carried out to forestall the possibility of German invasion and to hit at targets that supported the German air-sea blockade. They were carried out not only by Bomber Command but also by Coastal Command aircraft; in 1941, once the invasion threat had receded, Fighter Command also attacked coastal targets in large-scale fighter sweeps—the so-called Circus operations—planned to lure the German Air Force into combat and undermine air force organization. The strategic assault of economic and military targets nevertheless remained limited at first, partly from concern over the political implications, partly from the military risks of attacks in daylight (night bombing had not yet been approved for non-German targets) against the large German air forces stationed across northwest Europe. The reaction of the communities subjected to fringe bombing was mixed. There was evidence that the occupied peoples positively wanted the RAF to bomb the military and industrial targets in their midst. A Dutch request arrived in August 1940 to bomb the Fokker aircraft works in Amsterdam and a munitions plant at Hemburg (“working full capacity. Please bomb it”).⁸ A long letter from a French sympathizer forwarded to the Foreign Office in July 1941 claimed that many people in occupied France wanted the RAF to bomb factories working for the Germans: “The bombardments not only have a considerable material effect, but are of primary importance for the future morale of the anglophile population.”⁹ More letters arrived via Lisbon from Belgian sources explaining that the failure to bomb collaborating factories was attributed in Belgium to British “decadence.” A Belgian Resistance newspaper, Le Peuple, published a report of an informal referendum on bombing taken among workers in factories exploited by the Germans. “Not a single discordant voice,” ran the report. “They all wish for the destruction of plants which work for the enemy.”¹⁰

Alongside this more positive evidence, there were regular protests from the localities that were the object of the early fringe bombing and concern expressed by the governments-in-exile in London (Dutch, Belgian, Norwegian) as well as the Free French led by General Charles de Gaulle. The Dutch government-in-exile wanted assurances in 1940 that bombing would not harm Dutch civilians or civilian property.¹¹ The objections from French sources were a response to the regular small raiding that occurred throughout 1940 and 1941, a total according to French records of 210 raids in 1940 and 439 in 1941 in which 1,650 people were killed and 2,311 injured.¹² In May 1941 the mayors of the coastal towns of Dieppe, Brest, Lorient, and Bordeaux protested through the U.S. legation at Vichy about heavy bombing of residential areas. The British foreign secretary, Anthony Eden, asked the Air Ministry to take every care to minimize damage to civilian property and civilian casualties, but the raids continued nonetheless. Intelligence information suggested that the French population still believed that the RAF bombed only the military targets, while it was the Germans who bombed residential districts to try to stoke up popular hatred of the British.¹³ In August the Vichy regime made a formal diplomatic protest to the British government via the British embassy in Madrid about the inaccuracy of British bombing, followed by further representations in September about the continual bombing of the Channel port of Le Havre, where it was claimed that British aircraft had attacked the town fifty-five times in a year (it was indeed a regular RAF target), scattering bombs all over the residential quarters and killing 205 people. The municipal council of Le Havre recognized that the port was “on ‘the front’ in the war between Germany and England” but also pointed out that “no state of war between France and England exists.”¹⁴ The Air Ministry declined to reply to the French protest, but instead told the Foreign Office that accuracy was impossible when operations had to be carried out under indifferent weather conditions and the German habit of generating a smoke screen as soon as the bombers were sighted. The Bomber Directorate suspected that the protests were part of an orchestrated German plot to compel the RAF to reduce their offensive against French targets.¹⁵

A number of factors explain the escalating scale and lethality of bomber attacks on non-German targets from late 1941 onward. The military situation brought increasing pressure to bomb targets in occupied Europe that served the German submarine and air blockade. The prime targets were to be found in the ports of western France and the airfields and bases across the Channel in northern France and the Low Countries. It was also soon evident that armaments, aviation, and shipbuilding firms in the occupied zones were being utilized by the Germans, either directly taken over or the result of a collaborative agreement.¹⁶ The Ministry of Economic Warfare considered these to be necessary targets for air attack, not only because they were more accessible than most targets in Germany, but because their destruction might reduce the willingness of occupied populations to work for the extended German war effort. On June 23, 1941, the War Cabinet approved the bombing of factories throughout occupied France, but only by daylight, to ensure a better level of accuracy, though the RAF still held back on political grounds from bombing targets far inland, including Paris. Small raids on German shipping at Brest had begun in January 1941, but the first heavy raids against Brest and Lorient, including in this case night attacks, started later in the spring, though these too remained intermittent and ineffective until the War Cabinet recommended a sustained campaign in October 1941 to reduce the dangerous threat to the Atlantic sea lanes, which it failed to do.¹⁷ At the same cabinet meeting Churchill agreed to attacks on goods trains in northern France, which were assumed to be carrying supplies or ammunition for German forces. Step by step the military imperatives for bombing targets in occupied Europe pushed the RAF across the thresholds established in 1940.

The second factor was political. During the course of 1941, as it became clear that the war was to be a long-drawn-out conflict, the conduct of political warfare assumed a larger place. The Ministry of Economic Warfare, under the Labour politician Hugh Dalton, was at the heart of the indirect strategy laid down in 1940 to use bombing, blockade, propaganda, and subversion as the means to undermine German control of occupied Europe. In the summer of 1941 the minister of information, Brendan Bracken, proposed setting up a separate organization, the Political Warfare Executive (PWE), jointly run by the Foreign Office and the Ministries of Information and Economic Warfare, to coordinate the political initiatives directed at occupied Europe. It was formally constituted in the late summer under Robert Bruce Lockhart, with the journalist Ritchie Calder as director of plans and propaganda.¹⁸ The PWE directors immediately saw the connection between British bombing strategy and political propaganda. The Joint Planning Staff in June 1941 had already indicated that active armed resistance in Europe would never work “until bombing has created suitable conditions.”¹⁹ Calder began to lobby for a bombing policy governed not only by military and economic considerations, but by political calculation. He was impressed by the apparent enthusiasm for being bombed expressed in contacts with the occupied populations. The Norwegians in particular wanted to feel that they were still part of the war. Bombing, he thought, would “prove British interest in Norway”; Air Intelligence confirmed that Norwegians were “puzzled and bewildered” by the absence of raiding.²⁰ In a memorandum for the Foreign Office on the “RAF and Morale-Making,” Calder recommended bombing as a way to show the occupied populations that even if Britain could not invade, the German occupiers would not be immune from attack. On the other hand, he continued, “lack of British activity creates the impression that we have ‘abandoned’ the Occupied Territories.” Calder thought that “demonstration raids,” as he called them, would counter a mood of “listlessness and despair,” and invigorate militant forces among the occupied peoples.²¹

The weapon of political warfare was the leaflet rather than the bomb. Throughout the conflict the political warfare and intelligence establishment remained convinced that propaganda from the air was worthwhile, and millions of small sheets, or pamphlets or newsletters, were jettisoned over the target populations, both enemy and ally. Aircrews seem to have been less persuaded of the value, and the PWE directed some of its propaganda effort to instilling confidence among them that leaflets, or “nickels,” as they were known, were just another, and equally effective, tool in the Allies’ armory. “They are a weapon aimed not at men’s bodies,” ran one training manual, “but at their minds.”²² The task of drafting, translating, printing, and distributing the material was enormous. The statistics of RAF leaflet drops are set out in table 6.1; the wartime total was 1.4 billion.

Table 6.1: British Leaflet Distribution by Aircraft by Year and Territory, 1939–45 (in thousands)

Source: Calculated from TNA, FO 898/457, PWE, “Annual Dissemination of Leaflets by Aircraft and Balloon, 1939–1945.”

Each piece of aerial propaganda had to be discussed in terms of the current political and military situation, and the language adjusted accordingly. It also had to be considered that for many of those who risked picking up and reading the material, this was the only way they could get news of what was happening in the wider war. Allied confidence in the effects of leafleting was sustained by regular intelligence about the popular demand for more. In Belgium it was reported that children sold the leaflets they picked up for pocket money; French peasants concluded that if the RAF could waste time dropping leaflets, it “must be very strong.”²³ On the actual effect of leaflet drops the evidence remains speculative. In Germany and Italy it was a crime to pick them up at all.

There is little doubt that the PWE greatly exaggerated the political effects likely to be derived from a combination of propaganda and judicious bombing. Like the optimistic assessments of imminent social crisis in Germany in 1940 or 1941, every straw of information was eagerly clutched at. Violations of air-raid precautions were particularly highlighted. It was reported that seventeen Dutchmen had been heavily fined in the summer of 1941 for staying out on the street during a raid singing “Who’s Afraid of the Big Bad Wolf?” News from Denmark suggested that 20,672 prosecutions for blackout irregularities had been pursued in 1941.²⁴ British political warfare assumed that the working class would be the most likely to challenge the occupiers because they were by definition supposed to be antifascist. Directives to the BBC European Service in early 1942 asked broadcasters to “take absolutely for granted the workmen in enslaved countries are unhesitatingly behind our bombing policy, and will do all they can to help it.”²⁵ Bombing was supposed to suggest that liberation was close behind it and to encourage hatred of the German enemy. The leaflet campaign was deliberately designed to reflect this two-pronged argument. In the spring of 1941 messages to Belgium were to be divided into “Hope—45%,” “Hatred—40%,” “Self-interest—10%,” and “Self-respect—5%.” Propaganda aimed at the Netherlands had “Certainty of Allied Victory” top of the list, with 35 percent. In between the leaflets, the idea was to bomb intermittently to keep such hopes alive. In 1941 this appeal was possible. A Belgian woman who had escaped to Britain in October 1941 claimed the raids “were the best propaganda the British had done.”²⁶The years of apparent inactivity that followed undermined confidence in occupied Europe and dampened the hopes of Britain’s political warriors.

By the end of 1941 these military and political considerations combined to push the RAF toward a more vigorous and less discriminate bombing strategy for the occupied regions. At a War Cabinet discussion in November 1941, the air minister pressed for permission to begin nighttime raids against industrial targets across occupied Europe, including the major Renault works at Boulogne-Billancourt in Paris. Churchill insisted on postponing any decision until the political outcome was properly evaluated, but following RAF representations early in 1942, which claimed that the morale of the occupied population was better in areas that had been bombed than in those so far neglected, Churchill finally agreed to allow general bombing of European targets, and the cabinet confirmed the change at its meeting on February 5, 1942.²⁷ The RAF scarcely needed to be prompted. The Air Ministry in November 1941 had already discussed the use of incendiary bombs in attacks on industrial targets in occupied Europe to achieve maximum damage and “to gladden the hearts of all men and women loyal . . . to the Allied cause.”²⁸ In April 1942, Bomber Command was instructed to bomb targets in France, the Low Countries, and Denmark (“knock them about”), so that local people would demand proper protection and hence disperse the German antiaircraft defenses.²⁹ The PWE reached an agreement with the RAF to ensure that political considerations would play a part in target planning. The link between political propaganda and bombing policy became institutionalized and remained throughout the war a central element in bombing all the areas under German control.³⁰

France: Bombed into Freedom

The long arguments over whether or not to bomb targets in Paris were finally resolved by the decision in February 1942 to allow raids against important industrial targets throughout Europe. The raid on the Renault works became a test case of the dual strategy of economic attrition and morale making. On the night of March 3–4, Bomber Command sent off 235 bombers, the largest number yet for a single raid. Flying in to bomb from between 2,000 and 4,000 feet with no antiaircraft fire to distract them, 222 aircraft dropped 419 tons on the factory and the surrounding workers’ housing. Much of the factory area was destroyed, though not the machinery in the buildings, at the cost of only one aircraft lost. No alarm had sounded and casualties among the local population were high: French civil defense first reported 513 killed and more than 1,500 injured, but the Paris prefecture eventually confirmed 391 dead and 558 seriously injured, more than twice the number inflicted so far by the RAF on any one night over Germany. An estimated 300 buildings were destroyed and another 160 severely damaged.³¹

The works were bombed not only for the potential damage to German vehicle output in the plant, but also to test how French opinion might react to an escalation of the bombing war. Leaflets were dropped beforehand “To the populations of occupied France,” explaining that any factory working for the Germans would now be bombed and encouraging workers to get a job in the countryside or to go on strike for better protection; a BBC broadcast warned French people to stay away from collaborating businesses.³² The PWE wanted to find out as soon as possible after the raid how French workers had reacted, “because it is the workers who have been killed, the workers who ‘go slow’ and sabotage.”³³ Although the French authorities orchestrated elaborate public funerary events, the British soon received indications that the reaction had not been as adverse as the public outcry might have suggested. A report from Roosevelt’s special emissary in the new French capital at Vichy, William Leahy, explained that the propaganda campaign fostered by the regime with German support had been ineffective and that there was little evidence of anti-British feeling either in Paris or in the rest of unoccupied France. Anthony Eden, who had been anxious about the political effect, was pleased with the results of a “well-executed blow,” which he believed evoked “admiration and respect” among the people who suffered it. He was now willing to support further raids.³⁴ In Paris itself the operation was welcomed by many as a sign that liberation might be one step nearer. “Nobody was indignant,” wrote one witness. “Most hid their jubilation badly.” Blame was directed much more at the French and German authorities for failing to sound the alert, or to enforce the blackout effectively, or to provide adequate shelters.³⁵ Rumors quickly circulated outside Paris that the Germans had deliberately locked the workers inside the factory or had barred entry to the shelters. It was said that Parisians called out “Long Live Great Britain!” as they lay dying.³⁶ The raid itself had limited results. Reports reached London in June that only 10 percent of the machine tools had been lost as a result of the bombing and that the Renault works was operating at between 75 and 100 percent of its pre-raid capacity.³⁷

The heavy bombing of French targets between 1942 and 1944 by Bomber Command and the Eighth Air Force was undertaken in the hope that casualties could be kept to a minimum to avoid alienating the French population, while serious damage was done to Germany’s western war effort. It was unfortunate for the French people that heavy bombers were seen as the necessary weapon for a number of very different strategic purposes for which they were far from ideal.³⁸ From 1942 onward, bombers were used to try to destroy the German submarine presence on the French west coast by bombing the almost indestructible submarine pens and the surrounding port areas; in 1943–44 bombers were directed at small V-weapons sites that were difficult to find and to damage; in the months running up to the invasion of Normandy, the Transportation Plan similarly directed all Allied bomber forces (including the Fifteenth Air Force in Italy) against small rail targets, many of them embedded in urban areas; finally, the months of campaigning across France in the summer of 1944 led to regular calls from the ground forces for heavy bomber support, producing some of the most devastating raids of the war against French towns defended by German troops. The result was to strain popular French support for the bombing of the enemy in their midst.³⁹ Although anti-German sentiment was not reversed by the air campaign, there was a widespread belief that a less damaging strategy could have been found to achieve the same end.

The antisubmarine campaign exemplified the many contradictions that plagued the decisions to bomb France more heavily. When Bomber Command was directed to attack German naval targets on the French west coast, the orders were to attack only the dock areas and in conditions of good visibility. In April 1942, Harris wrote to Portal suggesting that the best way to slow down the German submarine war and to drive fear into the French workforce was to carry out “real blitzes” on Brest, Lorient, St.-Nazaire, La Rochelle, and Bordeaux.⁴⁰ Portal demurred since this was still contrary to government policy; in October 1942 guidelines were issued to Bomber Command to ensure that the air force would understand that only identifiable objectives in clear weather could be bombed and only if it was certain that heavy loss of civilian life would not result.⁴¹ But when the Atlantic battle reached its climax in late 1942, the Anti-U-Boat Committee, under pressure from the Admiralty, finally recommended abandoning all caution by destroying through area attacks the towns that involuntarily hosted the German submarines. The War Cabinet approved the decision on January 11, 1943, and although Harris by now no longer wanted operational distractions from his attacks on Germany, his desire for “real blitzes” on the French ports could now be fulfilled.⁴² Harris described the French interlude in his memoirs as “one of the most infuriating episodes” in the whole bomber offensive and an evident “misuse of air power.”⁴³ He blamed the Admiralty for the change in priority, and there is no doubt that the driving force behind it was the chief of the naval staff, Admiral Sir Dudley Pound, who in this case was able to persuade Churchill and Eden to swallow their scruples over bombing civilians for the sake of the survival of British sea traffic.

There had already been more than twenty attacks on the ports since 1940, which had served to encourage the Germans to take every precaution to protect submarine operations.⁴⁴ In the summer of 1942 a British propaganda campaign had been launched from the air against coastal towns from Dunkirk to St.-Nazaire warning the populations to evacuate: “we must carry on a war to the death [guerre à l’outrance] against the submarines.”⁴⁵ Most evacuations did not occur until the bombing started in earnest. The heaviest raiding was reserved for Lorient, where nine major attacks by Bomber Command in January and February dropped 4,286 tons of bombs (including 2,500 tons of incendiaries) with the specific purpose of burning down the town. On one raid the bombers carried 1,000 tons of bombs, the same quantity dropped a few months before by the German Air Force in the major raid on the city of Stalingrad.⁴⁶ The French report following the bombing described the raids as an example of a new RAF strategy of “scorched earth”; not a building in the town remained standing or unscathed, a “dead city,” except for the submarine pens undamaged by the rain of bombs. In a thirty-kilometer radius from the town thousands of village buildings had been destroyed and farms incinerated.⁴⁷ The PWE published an uncompromising statement following the bombing that the innocent must inevitably suffer with the guilty: “The violence and frequency of attacks involving hardship to civilians must increase.”⁴⁸ Naval Intelligence assessments were nevertheless unimpressed by Bomber Command’s strategy “of the bludgeon,” which failed to halt the rate of submarine operation significantly in any of the major targets, despite Pound’s earlier insistence that it would. In April, Harris was instructed to stop and to turn once again to Germany.⁴⁹ The submarine threat was defeated in spring 1943 by using aircraft to attack submarines at sea rather than in their concrete pens. The pens themselves became vulnerable only after the development of two giant bombs—“Tallboy” and “Grand Slam”—both the brainchild of Barnes Wallis, the engineer who designed the bomb used to breach the Ruhr dams. But the first five-ton Tallboy was only used against Brest on August 5, 1944, and Lorient a day later, while the ten-ton Grand Slam was available only for the last month of the war.⁵⁰

Bomber Command was joined by the Eighth Air Force for the submarine campaign, and the round-the-clock bombing gave the local population, most of whom had been evacuated or had sensibly evacuated themselves, no respite from the raiding. Daylight bombing was carried out from a considerable height by crews who were still learning their way. The wide spread of bombs dropped from high altitude and the rising casualty rates that resulted provoked a sudden change in French attitudes during the course of 1943. A French Resistance worker who arrived in Britain in April 1943 warned his new hosts that the population was deeply hostile to high-level American raids, which threatened to undermine irretrievably “the friendly feelings of the entire French population towards the Allies.”⁵¹ This shift in opinion coincided with the decision to spread the bombing over all French territory following the German occupation of the southern, unoccupied zone in November 1942. On December 21 the Air Ministry was informed by the Foreign Office that raids on southern French cities were now legally permitted, and on December 29 the BBC broadcast the same warning to the population living there to stay away from military and industrial targets that had been given to the occupied north earlier in the year.⁵² The guidelines issued in October 1942 on the conduct of raids now applied to the whole of France, but they were not binding on the Eighth Air Force, and when the Renault works in Boulogne-Billancourt were bombed again on April 4, 1943, by eighty-five B-17 Flying Fortresses, the results were very different from a year before. Just under half the bombs hit the industrial complex, but the rest were scattered over a wide residential area. One bomb penetrated the metro station at Pont-de-Sèvres; eighty corpses were identified there and the unidentified human remains put into twenty-six coffins.⁵³ There was little antiaircraft fire or fighter pursuit and only four aircraft were lost. The alarm had sounded only one minute before the bombs began to fall, giving the population out on the streets in the hour after lunch little chance to find shelter. The civil defense counted 403 dead and 600 injured; 118 buildings were destroyed and 480 heavily damaged.⁵⁴ Two days later, René Massigli, the French ambassador (representing the provisional government in London), met Eden to complain about the “feeling of exasperation” in France caused by civilian losses from careless American bombing. In Brittany, he claimed, the reaction of the population was to cry out, “Vive la R.A.F.!” but also “À bas l’Américain Air Force.”⁵⁵

Raids over the summer by both air forces were reined back. The Eighth Air Force was asked to confine raids just to the submarine bases and to try to find an operational pattern that would reduce French civilian deaths. Arnold objected to British requests to restrict what American air forces could do and an agreed list was drawn up of objectives in France that could be attacked after a warning to the population. Massigli was told by Eden that the American air forces would only bomb certain selected targets and would try to do so with greater care, but by the autumn Eaker was keen to extend the Pointblank attacks to aircraft industry targets in France.⁵⁶ The French aircraft industry, much of it sheltered in the unoccupied zone until November 1942, produced 668 aircraft for Germany in 1942, 1,285 in 1943, many of them trainer aircraft to free German factories for the production of combat models. German manufacturers used French capacity for their own experimental work, away from the threat of bombs on Germany.⁵⁷ As a result, French industry became a military priority for the American air force even at the risk of inflicting heavy casualties on the population. On September 3 and 15, 1943, Eighth Air Force raids on factories in Paris spread damage once again across residential streets packed with workers and shoppers and killed 377 civilians.⁵⁸ The raids on the western port city of Nantes on September 16 and 23 exacted the highest casualties so far from French bombing. The targets included a German naval vessel, a French locomotive works, and an aircraft factory, which was hit heavily. On September 16, 131 B-17s hit the town with 385 tons of bombs; on September 23, forty-six out of 117 B-17s dispatched to Nantes in the morning dropped a further 134 tons in poor weather, followed by a less accurate raid by thirty aircraft in the evening.⁵⁹ In the first raid the bomb pattern once again spread out over a wide area of the city, destroying 400 buildings and severely damaging another 600. The civil defense authorities counted 1,110 dead and 800 severely injured. While the local emergency services struggled to cope with the damage, they were hit by the two attacks on September 23, which not only struck the ruins but spread out over an area of more than 500 hectares. Because much of the center of the city was already abandoned, deaths from the second two raids were 172, but a further 300 buildings were completely destroyed. This time the population panicked entirely and 100,000 abandoned the city. The raids on Nantes resembled completely the pattern of raids on a German city, with the exception that Eighth Air Force losses were modest, a total of seven aircraft on September 16 and no losses a week later.⁶⁰

The raids of the autumn of 1943 provoked a mixture of outrage and incomprehension in France. Total deaths from bombing in 1943 reached 7,458, almost three times the level of 1942. A French report on public opinion, which reached the Allies early in 1944, highlighted the damaging effect of persistently inaccurate high-level bombing on a people “tired, worn out by all its miseries, all its privations, all its separations, unnerved by too prolonged a wait for its liberation.”⁶¹ The French Air Force, reduced under the armistice terms with Germany to a skeleton organization, tried to assess what object the Allied raids could have. Raids on Paris and against the Dunlop works at Montluçon (this time by Bomber Command) puzzled French airmen, who assumed there must be some secondary purpose behind the pattern of scattered bombing that they had not yet worked out.⁶² Since the French Air Force could not do its own bombing, much time was spent in 1943 and 1944 observing Allied practice in order to understand the techniques and tactics involved as well as the effects of bombs on urban society, industrial architecture, and popular morale.⁶³Many of the reports on individual raids highlighted the sheer squandering of resources involved in a bombing operation when three-quarters of the bombs typically missed the target: “The results obtained,” ran a report on the bombing of St.-Étienne, “have no relation to the means employed, and this bombardment represents, like all the others, a waste of matériel—without counting the unnecessary losses in human life that they provoke.”⁶⁴ The air force worked out the pattern of bombing accuracy to show just how wide the dispersion of effort was. In raids against Lille, the area in which bombs fell was a rectangle 8 by 4 kilometers; against Rouen, 8 by 3 kilometers; a raid on the railway station at Cambrai in 1944 covered an area 3 kilometers in length and 1.5 kilometers wide. The impact varied from raid to raid, but studies showed that many raids covered an area of between 200 and 400 hectares (500 to 1,200 acres), which explained the escalating losses of life and property. The French Air Force was impressed most by low-level dive-bombing and rocket attacks using the American P-47 Thunderbolt and the Hawker 1-B Typhoon, which achieved their object with much greater operational economy, and matched French strategic preferences before 1939.⁶⁵

The French government and population were not unprepared for a bombing war. As in Britain, the French state had begun to plan for passive defense against air attack as early as 1923; a law for compulsory passive defense organization was passed in April 1935, compelling local authorities to begin the organization of civil defense measures. In July 1938 a director of passive defense was appointed in the Defense Ministry to coordinate the protection of civilian lives and property with the committees of passive defense set up in each French administrative département.⁶⁶ The problem for French civil defense was the sudden defeat and occupation in the summer of 1940. In the area occupied by the Germans, civil defense was likely to be a necessary safeguard against British air activity; in the unoccupied zone, the urgency for continued civil defense seemed less evident. The Vichy government set up the Directorate of Passive Defense in the southern city of Lyon in 1941 under General Louis Sérant, but it was starved of funds and personnel. Spending on passive defense had totaled more than 1 billion francs in 1939 but by 1941 was down to just 250 million.⁶⁷ In both zones of France the difficult task was to reach a satisfactory working relationship with the German occupiers. The active air defense of the occupied zone was in the hands of Field Marshal Hugo Sperrle’s Air Fleet 3. Following the switch to the war against the Soviet Union, the number of fighter aircraft and antiaircraft guns left in France was seldom adequate for the weight of Allied attack. German priority was given to the protection of the most important military sites, including the submarine pens and German airbases. Air-raid alarms could only be activated on German orders, though French observers were expected to supply information to allow German officers to calculate whether it was worth sounding an alert. In the occupied zone the blackout was enforced on German orders. Mobile emergency units for air protection were sent from Germany to help with firefighting and rescue work alongside the residual French passive defense organization. They found the French attitude at times lackadaisical. German firemen fighting a blaze in Dunkirk in April 1942 were astonished at the lack of discipline among French colleagues who “stood around on the corners smoking.”⁶⁸

The relationship with the unoccupied zone was a constant source of friction for the German air command in Paris and the Italian occupation zone set up in 1940 in southeastern France. The Italian Armistice Commission insisted that Vichy impose a blackout throughout the area abutting the Italian-occupied regions to avoid giving British bombers an easy aid to navigation against Italian targets, but even when the French Air Force agreed, it proved difficult to enforce.⁶⁹ In November 1941 the German Armistice Commission in Wiesbaden complained that British aircraft regularly flew over the unoccupied zone without any blackout below: “The contrast between the occupied zone, plunged into darkness, and the unoccupied zone, where the blackout is up to now only intermittent, nicely indicates to enemy planes the frontier of the two zones.”⁷⁰ The German Air Force demanded complete blackout every night along a cordon 100 kilometers from the occupation zone, and effective blackout over the whole of unoccupied France when aircraft were sighted. French officials regarded the request as “inopportune” and prevaricated for months until August 1942, when the French government finally accepted a blackout of the frontier zones.⁷¹ A German aerial inspection a few weeks later showed that many houses had not bothered to take blackout measures; vehicles could be seen driving with full headlights; in Lyon the blackout occurred only after the antiaircraft artillery had begun to fire.⁷² The long delay reflected a more general reluctance on the part of the French military leadership to comply with German demands. Failure to observe the blackout was also a simple way to express noncompliance. Free French radio broadcasts encouraged householders to keep lights on throughout the night to help the RAF find German targets. Only when the whole of France was occupied could the German occupiers insist on the blackout, but even then complaints continued about its inadequacy.

The occupation of the southern zone on November 10, 1942, coincided with the intensification of Allied bombing. As this became heavier, the French authorities recognized that failure to collaborate fully with the German occupiers would expose the population to unnecessary risks. The ambivalence remained, however. When German Air Fleet 3 asked for French antiaircraft gunners in 1943 to man batteries in the north of the country, French officials preferred to site them in central France where they could be used for training purposes rather than to fire at Allied aircraft.⁷³ In February 1943 the German military command in Paris insisted that a unitary French antiaircraft defense system should be set up covering the newly occupied French territory and working in close collaboration with the thinly spread German antiaircraft resources. The Vichy regime was asked to establish a Secretariat for Air Defense, including a national director for “passive defense,” and it was the German intention that the French organization would eventually operate over the whole of France.⁷⁴ The new French defenses included antiaircraft batteries that were, unlike their German counterparts, controlled by the army. The German Air Force command in France insisted that the new French units come under air force control, and the army was forced to comply.⁷⁵ A new air-raid warning system, the Securité Aérienne Publique (SAP), was activated in February 1943, manned by French personnel under French Air Force control, using a mixture of radar and visual observation. In the southern zone the force numbered 3,800 officers and men; in the northern occupation zone the German Air Force still kept its own system of alerts, but Vichy officials and officers were posted to the main air defense centers to help coordinate air defense measures across the whole country.⁷⁶ The system suffered from the same problems found in the northern zone, since alerts could only be authorized by the Germans on information passed to them by French observers, except in more remote areas where there were no German officials.⁷⁷ The result once again was that alerts were sometimes sounded only when aircraft were already overhead, minutes before the bombs dropped.

The passive defense system insisted on by the German Air Force already existed in a skeleton form throughout Vichy France, organized by local prefects and mayors. In the southern zone the system had not been properly tested and now required a rapid expansion. The Vichy regime, now led by Pierre Laval as premier, established an Interministerial Protection Service against the Events of War (Service Interministériel de Protection contre les Événements de Guerre, or SIPEG), not unlike the committee established by Joseph Goebbels in Germany the same month, designed to oversee all the policies necessary to maintain economic and social survival in the bombed cities.⁷⁸ The Passive Defense Directorate, a branch of the new department of Aerial Defense, held an awkward constitutional position between the German authorities on the one hand and the French SIPEG on the other. One of the things the new organization had to provide was mobile support units to cope particularly with the threat of firebombing on the model already adopted in northern France. Emergency fire and rescue battalions were set up at Avignon, Lyon, Aix-en-Provence, and Montpellier, to be summoned with German approval to any raid where local civil defense could not cope.⁷⁹ But they remained short of personnel—there were only 1,500 to cover the whole of southern France—and short of essential equipment because it was being supplied by French factories to meet German orders. When units were sent to help with raids in northern France, the shortages of manpower and equipment were evident, while the population in the south complained that they were not left with adequate protection.⁸⁰ In general, French cities were much less well protected than British or German cities, while the tension between the French organization and the German authorities, whose principal interest was in safeguarding German military installations and industries working to German orders, left civilian communities potentially more vulnerable to the effects of inaccurate raiding.

In many cases, however, the German air defense forces cooperated with French civil defense and emergency services. At Lorient the German Air Protection Regiment 34, stationed in northwest France, was called in during January 1943 to try to stem the fires not only in the port area where German personnel were stationed, but also in the residential areas hit by the rain of incendiaries. The local civil defense also summoned help from seven fire services in other towns. The failure to save Lorient resulted not from the lack of effort on the part of both French and German emergency workers, but from the sheer weight of the attack.⁸¹ In Nantes, later in the year, the two forces, French and German, also cooperated in fighting the effects of the raid not only on the port, which the Germans needed, but also on the streets of the town itself.⁸² Again it was the scale of the bombing that made it difficult for civil defense to cope with the immediate crisis, but by the day following the heaviest raid, September 17, 1943, there were 800 French and German workers, helped by local miners and teams from the National Youth movement, opening roads, making damaged housing safe, and searching for buried survivors. Eventually 1,500 emergency workers and volunteers worked to restore some kind of order. They were hampered first by the lack of equipment—there were only four mechanical shovels and just fifty trucks—and then by the attacks that followed on September 23. On the following day only 400 men remained to tackle the rescue work, since many workers had fled with their families from the ruined housing. Eventually twice this number could be found, but the French authorities observed that many were German workers, who displayed a greater discipline because they had no personal ties to the city itself.⁸³

The disaster at Nantes highlighted the problem of orderly evacuation as a solution to the increased threat from bombing. Evacuation had always been the French state’s preference as a way of providing really effective salvation to the urban population, but after the disastrous results of the mass exodus in 1940 during the German attack, priority was given to trying to prevent extensive evacuation and to keeping families together. Here again the German occupiers played a central part in dictating the pattern of evacuations. Following the bombing of Lorient, in which thousands of workers and their families disappeared into the surrounding countryside, the German high command in Paris decided that the vulnerable coastal towns should be evacuated in a planned way, giving priority as in Germany to children, mothers, and the elderly. In Cherbourg the Germans demanded the evacuation of 30,000 out of the 50,000 inhabitants, in Dieppe and Le Havre around one-quarter of the population.⁸⁴ The evacuations were carried out despite the reluctance of many inhabitants to leave. In Cherbourg over one-third of the evacuees later returned in the summer and winter of 1943, while the German commanders were lobbied for permission for wives and young children to return to live with male workers regarded by the Germans as indispensable. Once Allied invasion in 1944 became likely, however, the German occupiers insisted that the populations of the northern littoral evacuate as fully as possible to avoid being in the battle zone. There were only 5,000 people left in Cherbourg when the American army arrived in June 1944.⁸⁵ The Germans insisted on similar measures on the south coast of France, where it was possible that the Allies might launch a surprise invasion. Since the cities of the south were also now threatened by heavy bombing, evacuation of the coastal zone was seen by the French government as a useful means to reduce casualties. Preliminary plans in January 1944 suggested the transfer of up to 485,000 people for whom transport and accommodation had to be found in inland rural areas unprepared for the exodus. The combination of bombing and imminent invasion forced the French government to produce coordinated plans to move their wartime refugees more successfully than in 1940.⁸⁶

Evacuation had already begun in 1942 on an improvised basis, and by early 1944 over 200,000 children had been moved from the most vulnerable cities. In December 1941 a scheme was established between the bombed city of Brest and the southern city of Lyon in which the bombed-out (sinistrés) were to be housed in Lyon and given welfare and funds by the council and population that adopted them. The scheme failed to attract even 100 children, since parents were reluctant to accept separation and the children were reluctant to go.⁸⁷ In 1942 other bombed towns either sought or were offered adoption by cities regarded as safe, including Le Havre, which was eventually adopted by Algiers, but much of the aid came in the form of money or clothes or books for the homeless rather than a new home. Most French evacuees moved to family or friends in nearby villages, and French planners insisted, against German objections, that on practical and political grounds it made more sense to house evacuees locally rather than in remote areas in central France. With the heavy bombing of Lorient, St.-Nazaire, and Brest in early 1943, the population flowed out into the surrounding countryside in tens of thousands.⁸⁸ On February 4, 1944, Laval issued comprehensive guidelines on evacuation policy following the severe bombing of the winter and the expectation that the military threat would escalate. The guiding principles of the program were the need for an ordered transfer of population and the consent of those to be transferred, “voluntary but organized.” The government favored persuasion using a program of posters, radio broadcasts, and public meetings. Priority was to be given to “the human capital of the Nation,” above all to children, who carried the demographic future of a postwar France.⁸⁹ Mothers and children and pregnant women were the chief categories, though the elderly and disabled were also included; those who remained were classified as “indispensable” (administrators and officials), “necessary” (laborers and white-collar workers, doctors, welfare workers), and “useful” (those who helped to maintain the activity of the indispensable and necessary). Families nevertheless remained unenthusiastic about evacuation; they feared looting if they left their homes, and disliked the loss of independence and reliance on welfare in the destination zones. Eventually around 1.2 million moved as refugees, evacuees, or bombed-out, most in reaction to the urgent imperative of survival.⁹⁰

It has sometimes been remarked that the French failed to exhibit the “Blitz spirit” evident in Britain, and later in Germany, in the face of bombing. In a great many ways the opposite is true. The French population faced an inescapable dilemma that made it difficult to know how to respond to the raids: they wanted the Allies who were bombing them to win, and they wanted the Germans who protected them to lose. Since they were not themselves at war, the sense that they represented a national “front line” against a barbarous enemy could not as easily be used to mobilize the population as it could in Britain and Germany. The bombing was not part of an orchestrated offensive against French morale, and civilians were not supposed to be a target; nor was bombing experienced either regularly or over a wide area, except for the bombing of northern France during the Allied invasion. French towns and cities were nevertheless caught between two dangerous forces, the German occupiers and Vichy collaborators on the one hand, and the Allied air forces (including the B-24 “Liberator”) on the other. Resisting the Germans by helping Allied aircrew or sabotaging what had not been bombed meant running the risk of discovery, torture, and execution that no one in Britain’s Blitz was expected to face. Lesser infractions—deliberate refusal to observe the blackout, or absenteeism from a civil defense unit—could be interpreted by the occupiers not simply as an act of negligence but as an act of resistance. When evacuees returned without authorization, the local German commanders withdrew ration cards or threatened the returnees with a labor camp. French people exposed to the bombs experienced double jeopardy, both the damage and deaths from raids and the harsh authority of the occupiers.

This dilemma was exploited politically by both sides during the war. The German propaganda apparatus presented the Allied air forces as terror flyers, as in Germany, and the French press was encouraged to focus on the barbarous and indiscriminate nature of the attacks. The Vichy authorities shared this perspective, and may indeed have believed it. Cinema newsreels on the bombing of French targets broadcast by France-Actualités carried titles such as “War on civilians,” “Wounded France,” and “The Calvary continues,” while after every major raid there were elaborate official funerals with full pageantry and speeches condemning the massacre of the innocents.⁹¹ Since the Vichy regime was widely unpopular among important sections of the urban population, the bombing was used as a way to show that the authorities cared about the welfare of the damaged communities and to forge links between state and people. The bombed-out were entitled to state welfare at fixed rates; the state paid the funeral expenses of bomb victims; evacuation costs could be met in full for transfers of less than fifteen kilometers’ distance; pensions were introduced for those disabled by the bombs, and for those widowed or orphaned in the raids.⁹² In addition, bomb victims were entitled to welfare assistance from two voluntary welfare organizations, the Secours National (National Assistance), reestablished in 1940 with Marshal Philippe Pétain as its president, and the Comité Ouvrier de Secours Immédiat (COSI, the Committee for Workers’ Emergency Assistance), set up following the Billancourt raid in 1942 under the collaborationist René Mesnard. Both relied on state funds as well as voluntary contributions, and both echoed the propaganda of the Vichy regime in condemning the bombing and highlighting the efforts to aid the victims as a means of binding together the national community. The COSI took funds directly from the German authorities and in reality distributed little of it to the bombed-out and much of it to the officials who ran it.⁹³ The committee did play a part in redistributing to the victims of bombing some of the Jewish apartments and furnishings confiscated under German supervision, while the money given to the committee by the Germans came from expropriated Jewish assets. The first consignment of Jewish-owned furniture was handed over to COSI in April 1942, and large quantities continued to be diverted to help the bombed-out until 1944, though an even greater volume was shipped directly to the Reich from France and the Low Countries to supply German civilian victims of bombing, a total of 735 trainloads during the course of the occupation.⁹⁴

The Allies, on the other hand, needed to present to the French population a clear justification for the bombing as the key to eventual liberation. This message worked well early in the war when there was hope that RAF raids signaled the possibility of an early invasion, but less well after years of waiting and in the face of rising casualties. The Allies tried to combine the bombing with direct support for the French Resistance, but at the same time to avoid operations that would undermine the credibility of resistance and push the French population toward grudging support for Vichy. Broadcasts from the BBC, which were widely listened to in France, encouraged the French population to see resistance and bombing as two sides of the same coin.⁹⁵ The leaflet war was designed to offer clear warnings to the areas scheduled for raiding as well as justification for attacks on German targets or collaborating businesses. Millions of propaganda notices and news reviews were dropped throughout the period, reaching a crescendo in 1943–44. The RAF dropped 155 million in 1942 and 294 million in 1943, the great majority from aircraft, some from balloons sent with the prevailing winds.⁹⁶ The Eighth Air Force began leafleting operations in late 1942 only after the initial effects of American raids had been assessed to see what kind of political message should be delivered. A special force of twelve B-17 and B-24 bombers was set up in 1943 tasked with distributing leaflets over the occupied territories as well as across Germany.⁹⁷ By February 1944 the Americans had dropped 41 million items, including the French-language paper America at War, which was used to explain the course of the conflict and the necessity for bombing French targets. In spring 1944 the quantity increased substantially to 130 million in March on the eve of the transportation campaign against French railways, and more than 100 million each month until D-Day. So heavy was the bombardment of paper that the German authorities in France organized leaflet squads with sharpened sticks to collect them before they were picked up by the local population.⁹⁸

The impact of the leaflet and broadcasting campaign was difficult for the Allies to assess since almost all the public media in Vichy France treated the bombing as an unmediated crime. Allied intelligence was faced with a barrage of information showing that the bombing was defined by its “terror character.” One newspaper, the Petit Parisien, following the bombing of Paris in September 1943 claimed that “the barbarians of the West are worthy allies of the barbarians in the East.”⁹⁹ The Mémorial de St. Étienne asked, “Will this destructive Sadism have no end? One is appalled before this mounting barbarity, this barbarity behind the mask of civilization.” The Allies recognized that the French reaction was not as simple as that, but there was increasing evidence that even among pro-Allied circles the mixed results of bombing raids provoked anxiety and hostility without at the same time undermining the acknowledgment that German targets were both legitimate and necessary.¹⁰⁰ This ambiguity was evident from the reaction to two raids on Toulon on March 7, 1944. The first killed or injured an estimated 900 German soldiers and won wide approval; the second four days later missed the target and killed 110 French civilians to widespread complaints. One of the American crews shot down on the second raid was black, prompting racist comments about the quality and competence of American airmen.¹⁰¹ American bombing was identified in information from the French Resistance as the major source of resentment because of its apparently “careless and casual” attitude to the communities being bombed: “The Americans make it a sport,” ran one report, “and amuse themselves by bombing from such altitudes.”¹⁰² In May 1944 the French Catholic cardinals sent an appeal to the Catholic episcopate in Britain and the United States asking them to lobby the air forces to bomb military objectives with greater care and avoid the “humble dwellings of women and children.” The archbishop of Westminster replied that his government had given every assurance that casualties would be kept to a minimum.¹⁰³

There was nevertheless widespread resistance or noncompliance prompted by the bombing campaign as well. The Resistance took the view that those killed in Allied bombings were in some sense not victims, but combatants in a war for the liberation and salvation of the nation.¹⁰⁴ Those who chose to operate networks for the escape of Allied airmen certainly ran the risks of any combatant if they were caught. The death penalty was introduced in a decree on July 14, 1941, for helping Allied airmen, but an estimated 2,000–3,000 British and American servicemen were smuggled out of France and back to combat. In cases following heavy bombing, as at Lorient in 1943, some airmen were surrendered to the Germans, but Allied intelligence found that in many cases the Resistance distinguished between the regrettable effects of a heavy bombardment and their view of Allied aircrew as liberators.¹⁰⁵ The Resistance also regarded bombing as complementary to forms of active opposition to the occupiers, though it was seldom integrated as closely as it could have been, despite the insistence of the Resistance that sabotage could often be a more effective tool than bombing.¹⁰⁶ There also existed many lesser levels of protest or noncompliance derived from the bombing war. The funerals of Allied aircrew killed in action attracted large crowds despite German efforts to obstruct them; wreaths were laid by the graveside dedicated “To Our Heroes” or “To Our Allies” or “To Our Liberators” until seized or destroyed by the occupiers. There were numerous public demonstrations under the occupation, 753 in total, some orchestrated by Vichy to protest against bombing, but hundreds directed at shortages of food or adequate shelter.¹⁰⁷ The police reports from the provinces in 1943 found that despite, or because of, the bombings, the population talked openly of their hope for Allied invasion and the horrors of occupation: “No one,” ran a report from Charente in northwest France, “believes any longer in a German victory.”¹⁰⁸

The German occupiers found regular evidence of dissent among the French officials and servicemen organizing the air defense system. The slow introduction of French antiaircraft units in the summer of 1943 was blamed by the German Air Force on the existence of a network of Freemasons among the French officials involved. French antiaircraft personnel were made to sign a “declaration of duty” not to reveal military secrets, and both antiaircraft units and the French emergency services were monitored by the German Security Service (SD) for their alleged sympathies with de Gaulle and the Free French.¹⁰⁹ In August 1943, fifteen antiaircraft servicemen abandoned their posts and could not be found; the following month another fifteen men from the Air Force Security School took two cars and a truck and absconded to the Massif Central to join the partisan Resistance. In November 1943 a group of SAP soldiers were caught listening to French broadcasts from Britain; on the wall of their common room a poster was found proclaiming, “Vive les Gaullistes! Vive l’U.R.S.S.! Vive de Gaulle!”¹¹⁰ German Air Intelligence found that by the autumn of 1943, Allied success in the Mediterranean had changed the attitude of the French population to one of anxious longing for the moment of Allied invasion and celebration of every German defeat. “The expected Anglo-American landing in France,” concluded a report in August, “is now the daily topic of conversation.”¹¹¹

Allied planning for the liberation of France was indeed far advanced by the autumn of 1943, but from the Allied point of view it was bound to cause high casualties and perhaps compromise at the last moment the sympathies of the French people for the Allied cause. Churchill remained continually anxious, as he told the War Cabinet in April 1944, that preinvasion bombing might create an “unhealable breach” between France and the Western Allies.¹¹² The principal issue was the decision to use the heavy bomber forces, including the Fifteenth Air Force in Italy, to attack the French transport system before invasion and to support the army as it consolidated its position on the bridgeheads in Normandy in June and July 1944 and, a month later, in southern France. To this was added the decision to use Allied bomber forces in the Crossbow operations against German V-weapon sites across northern France. Neither Harris nor Spaatz was enthusiastic about using the bomber force this way, since it was not what the bombing was supposed to be for, while the aircraft had not been designed for use against small tactical targets. In January 1944, following an order to intensify raids on V-weapon sites, Harris rejected the use of Bomber Command to attack Crossbow targets as “not reasonable operations of war.”¹¹³ His reaction to the idea that bombers should support the ground offensive was just as negative. Bombers used for ground support would, he argued, “be entirely ineffective,” leading “directly to disaster” for the invasion force.¹¹⁴ Spaatz objected to Eisenhower that support for invasion was “an uneconomical use” of the heavy bomber force and preferred to leave the operations to the large tactical air forces assigned to the Allied Expeditionary Air Force under Air Marshal Trafford Leigh-Mallory, whose fighter-bomber and light bomber aircraft were intended to attack small targets and could react quickly and flexibly to battlefield requirements.¹¹⁵ Both bomber forces wished to be able to concentrate on Pointblank operations against Germany as a more strategically valuable way to limit the German response to invasion. Arnold told Spaatz in late April 1944, after the decision had already been taken to focus on the French railway system, that Pointblank should “still be pressed to the limit.”¹¹⁶ The arguments put forward in favor of the Transportation Plan by Tedder and his scientific adviser, Solly Zuckerman, have already been discussed. Zuckerman’s paper produced in January 1944 on “Delay and Disorganisation of Enemy Movement by Rail” formed the basis of the eventual preinvasion plan. On March 25 in a long and hotly debated meeting, Eisenhower finally came down in favor of using the bomber forces, under his own direct command, to attack the French railway system and other strategic targets both before and during the invasion period.¹¹⁷

This decision still left unresolved the political anxieties about possible levels of casualty. Portal informed Churchill after the meeting of March 25 that there were bound to be very heavy casualties as a result of the decision to hit seventy-six key points in the French railway network. Bomber Command suggested a figure of between 80,000 and 160,000 casualties, partly to confirm Harris’s argument that heavy bombers were the wrong weapon.¹¹⁸ Zuckerman calculated on the basis of damage done to British targets earlier in the war a more modest casualty figure of 12,000 dead and 6,000 seriously injured. In a discussion with the Defence Committee on April 5, Churchill deplored a strategy that might result in “the butchery of large numbers of helpless French people,” but despite his reservations and the opposition of Eden and General Brooke, chief of the general staff, the campaign was allowed to start on the understanding that casualty levels would be carefully monitored over the weeks that followed and warnings sent to French communities to evacuate the threatened areas.¹¹⁹ By mid-April casualties from the first nine raids were estimated at 1,103, well within the limits set by Zuckerman’s estimate. The Defence Committee was supplied with the outraged French reports (“In Anglo-American eyes, to be European is enough to be wiped off the list of the living”), and Churchill hesitated to give the campaign full approval.¹²⁰Zuckerman and the RE8 department of the Ministry of Home Security continued to monitor reports on a daily basis, and by late April the available evidence suggested that casualties had been approximately 50 percent lower than anticipated.¹²¹ Only after Roosevelt had insisted that there should be no restriction on military action if Operation Overlord were to succeed did Churchill finally on May 11 give his full approval to the campaign.¹²² For the four weeks before D-Day a furious crescendo of bombing descended on the French railway system and the unfortunate housing that surrounded its nodal points.

Zuckerman’s calculations in fact underestimated French casualties by a wide margin because transport targets were only part of what Allied air forces were expected to bomb in the weeks leading to invasion. French civil defense officials counted 712 dead in March, 5,144 in April, 9,893 in May, and an estimated 9,517 in June. The total of 25,266 over the four months was almost certainly not complete, given the difficulty of constructing exact records in a dangerous war zone; nor did all the casualties come from attacks on rail targets, but also against bridges and military installations, and German forces.¹²³ They nevertheless represented the overall human cost of the decision to use bombing as the means to reduce the capacity of the German army and air force to oppose the landings in Normandy. The high casualties resulted chiefly from the wide dispersion of bombs against relatively small targets and the large tonnage employed. The 63,636 tons dropped on transport targets exceeded the entire tonnage dropped by the German Air Force during the Blitz on Britain. The French air defense counted 71,000 high-explosive bombs between January and March 1944, but 291,000 from April to June.¹²⁴ Some attacks achieved a high level of precision, but in many cases bombs were scattered over a wide area. The attack on the rail center at St.-Pierre-des-Corps on April 11 struck the whole area of the town; the raid on Lille on April 10 hit an area of thirty-two square kilometers; that on Noisy-le-Sec on April 18 covered thirty square kilometers; on Rouen a day later, the area was twenty-four square kilometers.¹²⁵ In May the French authorities counted a total of 1,284 raids in which bombs fell on 793 different localities, 630 of them along the northern coast and the area northeast of Paris. Only 8 percent of the attacks were undertaken at night, which ought to have increased the possibility of more accurate raiding, but many of the daylight raids were carried out at heights of 3,000–4,000 meters (10,000–13,000 feet). In some cases, high casualties resulted from what the Passive Defense Directorate called “imprudence”—people standing at their windows to watch the bombing, others out in the street, or in their gardens. In a raid on Nice on May 26, 438 people were killed, two-thirds of them on the street, one-third in their houses. The shelters, for the most part either trenches or converted cellars, had uneven fortunes during the raiding; some stood up well even to direct hits, others, like one at Rouen on May 30, were blown apart, and most of the occupants killed.¹²⁶

Some of the heaviest losses of life occurred in targets in the former unoccupied zone, which were hit by American aircraft of the Fifteenth Air Force operating from bases in Italy. For the crews involved, the bombing of precise railway targets with a view to reducing damage to civilian lives and property was very different from the long-range raids against Pointblank targets in southern Germany, which had been the main activity of the force since its formation in November 1943. Two raids, one on St.-Étienne on May 26, 1944, and one on Marseille the following day, resulted in heavy loss of civilian life. At St.-Étienne the alert sounded in good time; the 150 B-17s attacked in waves from around 13,000 feet, and half the bombs fell in the zone around the rail links. But there were too few proper shelters for a population unused to the air threat and more than 1,084 were killed. The effect on rail traffic was limited. Rail lines remained open and the damage, such as it was, could be overcome in just four days. The attack on Marseille on May 27, flown at an estimated 20,000 feet, against stations at St.-Charles and Blancarde, both situated in the heart of the city’s residential area, scattered bombs over ten of the city’s quarters, destroying 500 buildings and killing 1,752 people. Again Passive Defense observed the “insouciance” of a population hit by an air raid for the first time and the absence of effective civil defense training. The stations were unimportant (one was a railway cul-de-sac), but the effect of the raid was to create a crisis of public morale and strong hostility to the air forces that carried out the attack.¹²⁷ The scale of the raiding and the damage inflicted brought protests from the French Resistance and the French authorities in London. The French Commissariat for Foreign Affairs warned the Foreign Office in early May that the raids were having a damaging effect on French opinion; in early June a resolution from the Resistance Group Assembly was passed on by Massigli, calling on the bomber forces to change their tactics and for an active propaganda campaign “to dissipate the growing ill-feeling” among the victim populations.¹²⁸ An OSS report from Madrid relayed the Resistance view that the French population now believed its situation to be no better than that of “the Nazis in Germany.”¹²⁹ This knowledge made little difference to Allied operations. In June, however, the bombing reached its high point as Allied forces poured ashore on D-Day and spread out into the Normandy countryside.

The results of the Transportation Plan were the subject of keen argument both at the time and since. French investigations showed that by the beginning of June rail traffic was down to around half the level in January 1944, and in the key regions of the north and west, down to 15 and 10 percent. There were 2,234 cases of damage to rail lines between January and June 1944, but as in Britain or Germany or the Soviet Union, these were relatively easy to repair.¹³⁰ Much damage was also done by sabotage, which the Resistance thought was a more effective way of achieving the same end, and with fewer losses to the French population, particularly the railwaymen, who were regarded as key Resistance workers.¹³¹ Between January and July, bombing and strafing destroyed or severely damaged 2,536 French locomotives, sabotage a further 1,605. But according to the SNCF (the French national railway), sabotage accounted for 70,000 goods wagons compared with 55,000 from air attack.¹³² In the three months from April to June there were 1,020 bomb attacks on the rail network, but 1,713 acts of sabotage.¹³³ Of these the two most significant causes of delay to traffic were the attacks on repair depots, which created a cumulative backlog of repair to the rolling stock hit by raids or sabotage, and the attacks on rail bridges. Many of these were carried out by the tactical air forces using fighter-bombers and light bombers, and they proved decisive in cutting the key regions off from rapid German reinforcement. Most rail centers could be made operable again in an average of seven days, but bridges took from ten to sixteen days.¹³⁴ The German authorities made strenuous efforts to keep the rail system going and succeeded for much of the period of the transport campaign. By suspending almost all civilian traffic and helped by persistent poor weather for bombing, it proved possible to maintain military through traffic up to June (when 535 loaded troop trains could still be deployed), but a slow decline set in from July. Total German ton-kilometers were 300 million for the month to mid-March, 400 million for each of the next three months, but only 150 million in July, by which time the loss of rail traffic compromised the further possibility of effective German defense.¹³⁵ The argument from the French viewpoint, however, was not whether German fighting power was affected, but whether the high cost in civilian lives and buildings could not have been avoided by wielding an aerial weapon that was less blunt. French authorities found that major raids by heavy bombers placed between half and four-fifths of the bombs outside the target area; in this sense Harris and Spaatz had been right to insist that large formations of heavy bombers were not the most suitable means to achieve the aim of precise destruction and limited French losses.

This conclusion was even more evident in the efforts of the two bomber forces to destroy the sites from which V-weapons were to be fired rather than raid the factories where they were being made. The first raids against the construction sites and depots in France were made in November 1943 after the Central Interpretation Unit at Medmenham had identified the first V-1 bunkers. The campaign against the V-weapons was code-named Crossbow, but the bombing operations were known as Noball. The quantities of bomb tonnage dropped during the course of the campaign, from early December 1943 to mid-September 1944, exceeded by a wide margin the total devoted to the Transportation Plan, a final tally of 118,000 tons of bombs, 86,000 of which were dropped between June 12 and September 12, 1944, on targets considerably smaller than the marshaling yards and viaducts targeted for D-Day. The first bombings in the winter of 1943–44 were thought to have set back the onset of the V-weapons campaign by six months, but after the first attacks the Germans abandoned the system of “ski-jump” launch sites (so called after their shape) because of their visibility and vulnerability, but let the impression remain that work was still being done on them in order to attract the bombers.¹³⁶ Eventually most of the original sites were identified and destroyed, but the newly modified launch sites were hard to find or hit. The German campaign was held up chiefly because of technical problems in producing sufficient operational V-1s to be able to start the offensive sooner.¹³⁷ After the first V-1s fell on London from the middle of June 1944 onward, a renewed order went out to both bomber forces to try to stamp out the threat. From December 1943 to May 1944, Crossbow targets had taken 12 percent of the bombing effort, but between June and August 1944 the proportion was 33 percent.¹³⁸ This represented a very large diversion of resources from any assistance that could be given to the Allied armies in France against targets that were almost immune to bomb attack. In April 1944, RE8 had explained to the Air Ministry that small sites protected by twenty feet of concrete had a low level of vulnerability.¹³⁹ In July, Sinclair instructed Portal to give the Crossbow sites a lower priority because “they are hard to destroy and easy to repair.” When Eighth Air Force B-17s attacked ten sites in July, they missed eight and dropped only four bombs on the remaining two.¹⁴⁰ Although Churchill had been keen for Bomber Command to try to blunt the V-weapon assault, the Air Ministry recognized by July that any effects were likely to be ephemeral. An Air Intelligence report in July on the V-1 sites captured by the American army in Cherbourg showed that although they had been heavily bombed, the design of the sites made them almost impervious to bomb damage and easily repaired if a chance hit achieved anything.¹⁴¹ The bombing continued until September, when most of the sites were captured by the advancing army, but both air forces recognized the limitation of using heavy bombers for what were in effect tactical targets.

The same limitations operated with the decision to use heavy bombers in support of Eisenhower’s ground campaign in France. For almost three months, northern France was a battlefield. As in the German attack on France in 1940 or the Soviet Union in 1941, it proved very difficult for the advancing Allied armies and air forces to avoid heavy damage to the towns, cities, and civilians in their path. In northern Normandy, where the battle lasted longest and was at its most intense, 14,000 French civilians died, 57 percent as a result of bombing. Heavy air raids began from the first morning, June 6, after warning leaflets had been dropped at dawn encouraging the Normandy population to “Leave for the Fields! You Haven’t a Minute to Lose!” In Caen on June 6 around 600 were killed by an American air raid, another 200 the following day amid the ruins of much of the city; on June 7 a raid by more than 1,000 Bomber Command aircraft against six small towns, including Vire, St.-Lô, Lisieux, and Coutances, eradicated the urban areas almost entirely. In the first two days of the campaign, 3,000 French civilians were killed.¹⁴² The village of Aunay-sur-Odon, bombed to stop the movement of German tanks a few days later, was literally erased from the map. Pictures taken after the raid showed a single church spire in an otherwise entirely level landscape. The French authorities counted 2,307 bombardments in June, 1,016 of them on the north coast provinces, most against railway targets. In July there were fewer raids, 1,195 in total, in August 1,121.¹⁴³ The great majority of the raids were tactical, carried out by the Allied Expeditionary Air Force, but on occasion the heavy bombers were asked to bring overwhelming firepower to bear. Two attacks on Caen, one on July 7 and a second on July 18, were among the heaviest of the Overlord campaign. The raid on July 7 involved 467 bombers dropping 2,276 tons on the northern outskirts of the town. There were few German defenders and the main effect of the raid, which left a moonscape on the approaches to Caen, was to force the British and Canadian troops to clear the roads before any further advance could be made.¹⁴⁴ The raid on July 18 by 942 bombers dropped an extraordinary 6,800 tons on the city and its eastern environs; the result did little to the German defenders, who had largely withdrawn to a defensive line south of Caen, nor to the population, 12,000 of whom eked out a precarious existence in caves outside the town at Fleury, but the raid once again left a ruined landscape that slowed down the advance of ground forces. By the end of the invasion a combination of bombing and shelling had left habitable housing for only 8,000 out of the 60,000 people who had lived there.¹⁴⁵

The weight of attack that could now be employed by the bomber commands was out of all proportion to the nature of the ground threat and on balance did little to speed up the course of the campaign. The establishment of air superiority over the battlefield was assured by the thousands of fighters and fighter-bombers available to Leigh-Mallory to establish a protective air umbrella over the Allied armies. Occasionally the bluntness of the bombing weapon spilled over to impose friendly fire on Allied troops. On July 24, on the eve of the American breakout into Brittany, code-named Operation Cobra, hundreds of Eighth Air Force bombers were ordered to shatter the German defenses in front of General Omar Bradley’s armies. “Ground grunted and heaved as the first cascade of bombs came down,” wrote Captain Chester Hanson in Bradley’s war diary, “horrible noise and the shuddering thunder that makes the sound of the bomb so different from the artillery.” It was followed by the sight of ambulances streaming to the front line to pick up the dead and injured from among the American troops hit by the bombardment, a total of twenty-five killed and 131 wounded. Among the victims was Lieutenant General Lesley McNair, whose mangled body was thrown sixty feet by a bomb and could only be identified by the three stars on his collar.¹⁴⁶ More bombs fell on American troops the following day, bringing the total dead to 101. Eisenhower decided not to use heavy bombers again to support the ground battle but to use them against targets he properly regarded as strategic, but Bradley once again called in heavy bombers to help unblock German opposition in Aachen in November 1944.¹⁴⁷ This time elaborate precautions were taken to ensure that the 2,400 American and British bombers used did not impose friendly fire on American forces. Large panels visible from the air were used as checkpoints in Allied lines to indicate clearly where the army was; a line of vertical radar beams was then set up by mobile units that could be distinguished by onboard radar in the approaching bombers; barrage balloons with special cerise markings flew at 1,500 feet in front of the American line, and antiaircraft guns were set up to fire colored flares at 2,000 feet below the bombers. Despite the most elaborate of precautions, two bomb batches still fell on American troops, but with only one casualty.¹⁴⁸ Aachen was turned into a wasteland.

The gulf separating means and ends in the application of heavy bombers to the campaign in France was no more evident than in the fate of two coastal towns that were obliterated by the Allied bomber commands. Both towns held stubborn German garrisons that refused to surrender even when all France had been liberated. The Channel port of Le Havre, subject to 153 small attacks since 1940, was no stranger to bombardment. It was strategically important as a potential port for Allied supply as Eisenhower’s armies moved rapidly eastward toward Germany, but it was defended by a garrison of over 11,000 German troops commanded by Colonel Eberhard Wildermuth. Since Le Havre was heavily fortified and he was under orders to prevent the port falling to the enemy for as long as possible, he rejected a request to surrender on September 3. Bomber Command was then ordered to bombard the city for a week before a ground assault could finally seize the port. A remarkable 9,631 tons of bombs were dropped and 82 percent of the town was destroyed at a cost of at least 1,536 civilian deaths. The German command refused to give up and a brief ground assault soon captured the port and the entire garrison. The post-raid analysis carried out by SHAEF concluded that the bombing had not done much to assist the eventual ground assault, a view that Harris shared.¹⁴⁹ Wildermuth cited artillery as the real source of the Allies’ rapid success on the ground; bombing killed only a tiny handful of German soldiers.

The second port was Royan at the mouth of the river Gironde, where the garrison had also refused to surrender when the whole surrounding area had been liberated. The presence of German forces made it difficult for the Allies to use the neighboring port of Bordeaux, and in December 1944, SHAEF was requested by the local American army commander to lay on a heavy bombing to push the garrison to abandon the fight. On the night of January 4–5, 1945, 347 Lancasters dropped 1,576 tons, including 285 4,000-pound “blockbusters”; around 85 percent of the town was destroyed and 490 French civilians (and 47 German soldiers) were killed. Poor communications had failed to alert Harris to the fact that targets outside the town had in fact been requested, not the town itself, while the French authorities had insisted that the civilian population had already been evacuated, which was not true.¹⁵⁰ The raid achieved nothing. The German commander refused to surrender until two further attacks by the Eighth Air Force on April 14 and 15, which dropped another 5,555 tons, destroying everything still standing. The two raids, one of 1,133 bombers, the second with 1,278, were the largest operations mounted against any target in France. The Germans surrendered three days later. A French journalist “defied anyone to find even a single blade of grass.”¹⁵¹

The overall cost from bombing in French lives and property during the war was exceptionally high, and it resulted in the main from using the overwhelming power of the bomber forces against modest targets that might more easily have been attacked by tactical air forces with greater accuracy. It was this lack of proportionality that attracted most criticism from French sources sympathetic to the Allies and eager that German targets should be bombed. “That which revolts the vast majority,” ran an intelligence report from December 1943, “of whom a great number are members of the Resistance is the inaccuracy of aim.”¹⁵² The result of using large heavy bomber forces in level flying from high altitude was to exact forms of damage not very different from the impact on German targets. Table 6.2 shows the overall cost of the bombing on France. The official figures presented here are lower than the figure of 67,000 for overall deaths regularly cited in the postwar literature, and the Passive Defense authorities regarded the initial statistics as a minimum. But although there are minor discrepancies in the figures published by different agencies in 1945, and a more general problem in classifying deaths caused by bombing, tactical air raids, or artillery fire in a battle zone, a figure between 53,000 and 54,000 dead is unlikely to be superseded by anything more precise.¹⁵³ The figure for 1940 includes deaths and destruction from the air inflicted by all air forces during the German invasion in May and June.

Table 6.2: French Losses from Bombing, 1940–45

Source: BN, Bulletin d’Information de la Défense Passive, May 1945, 4.

In the face of the political anxieties regularly expressed in London, why did the Allies use the bomber force in France with such apparent disregard for civilian losses? The bomber commanders were themselves unhappy with what was being asked of them. Spaatz considered the tactical air forces adequate for giving effective ground support. In notes for Eisenhower he argued that strategic bombers would not yield a sufficient strategic return if used for the invasion: “The advantages gained by such use would be very small compared to the effort put forth.”¹⁵⁴ The Eighth Air Force commander, Jimmy Doolittle, told Eisenhower and Spaatz in August that the use of strategic bombers with insufficient training and planning time in support of ground operations was bound to produce errors in execution and admitted that “the fighters have done a better job of supporting the Army than the bombers.”¹⁵⁵ The persistent use of the strategic forces has a number of explanations. For the Allied Supreme Command in Europe and the Combined Chiefs of Staff, bombing had evident advantages: it would speed up the invasion after years in which demands for the second front had failed to materialize; it would help to bring the end of the war closer for democratic populations anxious that hostilities should end sooner rather than later; it would make victory in France more certain and less hazardous; and it would finally allow the Allies to cash in the very large investment already made in strategic bombing that had not yet delivered what had been promised from the Combined Bomber Offensive. The bomber commanders also bear some of the responsibility. By making repeated and often strident claims about the capacity of strategic bombing to make a decisive contribution to shortening the war, they invited the ground forces to exploit those claims in a campaign that was regarded as decisive for the war effort. In the end, the balance between operational and political calculation was bound to fall in favor of the anticipated military outcome. When Eaker asked Portal in 1943, before starting the heavy bombing of French targets, how to overcome political objections to French losses, Portal replied that the government “have never shrunk from loss of civilian life where this can be shown to be an inevitable consequence of a considered and agreed plan.”¹⁵⁶

Eastern Europe: Everywhere but Auschwitz

The first time the RAF was invited to bomb the camp at Auschwitz () in Poland was in January 1941. At that time it was not the extermination and labor complex of Auschwitz-Birkenau, where more than a million European Jews were murdered between the spring of 1942 and the end of the war, but a camp for 20,000 Polish prisoners of war. The request came, according to General Sikorski’s Polish army headquarters in Britain, from the prisoners themselves, who would welcome a bombing raid that would allow them to escape en masse. Air Marshal Peirse, commander in chief of Bomber Command, replied that it was impossible. On clear nights every bomber had to be deployed against German industry. The German war economy, Peirse explained, was likely to experience a crisis in 1941. “Sporadic attacks” against a target such as Auschwitz were unlikely to be accurate enough to do more than kill many of the prisoners.¹⁵⁷

The next time the RAF was asked to bomb Auschwitz was in July 1944, when it was no longer a prisoner-of-war camp, but the center for the mass murder of European Jews. The complex consisted of three main areas: an extermination camp at Auschwitz-Birkenau; a camp for forced labor selected from those deemed fit enough from among the arrivals; and an industrial complex at nearby Monowitz where the chemical giant I.G. Farben was constructing a plant to produce synthetic rubber and other war-related chemicals. On July 7, following an interview with Chaim Weizmann, president of the Jewish Agency, Eden wrote to Sinclair asking whether it was possible to bomb the camp or the rail tracks leading to it. Churchill was keen to pursue this, but Sinclair, like Peirse, was unsympathetic. He told Eden that interrupting rail traffic in France had proved difficult even with the full weight of Bomber Command behind it; to find and cut a single line far away in Poland was beyond the power of the bomber force. Sinclair doubted that bombing the camp or dropping arms to the prisoners “would really help the victims.” He thought the American air forces might be in a better position to do it, and promised to discuss the issue with Spaatz, overall commander of American air forces in Europe.¹⁵⁸ Spaatz was sympathetic, but claimed that nothing could be done without better photographic intelligence of the camp itself. There was extensive reconnaissance material on the nearby Monowitz plant and other war-economic targets around Auschwitz, but although some photographs showed areas of the camp, the extermination center had not been the object of a specific reconnaissance operation.¹⁵⁹ Unknown to Spaatz, the War Department in Washington had already been lobbied several times in the summer of 1944 to undertake bombing of the rail lines or the gas chambers but had deemed the operation to be “impracticable.” On August 14 the assistant secretary of war, John McCloy, rejected the request (and did so again when lobbied in November).¹⁶⁰ Two weeks later the Foreign Office informed Sinclair that since the deportation of Hungarian Jews to Auschwitz-Birkenau appeared to have been halted, there was no longer any need to consider an operation to bomb it. On September 1, 1944, Spaatz was instructed to pursue the idea no further.¹⁶¹

There has been much academic argument over the question of whether an operation against the Birkenau extermination facility or the railway lines was feasible or not.¹⁶² There is no doubt that had it been a priority target for the Allies, it certainly could have been bombed. At just the time that the Allies were considering the requests from the Jewish Agency to undertake the bombing, the U.S. Fifteenth Air Force began a series of raids on the I.G. Farben complex at Monowitz, where the prisoners in the Auschwitz labor camp were marched to work every day. Auschwitz had been on the Mediterranean Allied Air Forces’ target list since at least December 1943, when plans were drawn up for attacks on German oil and chemical plants in eastern Europe.¹⁶³ The first raid on August 20, 1944, hit Monowitz accurately, a second on September 13 was hampered by enhanced German antiaircraft defenses, the third and fourth attacks on December 18 and 26 did more damage to the plant, and it was finally abandoned in January 1945 as the Red Army drew near. The damage was not extensive and output of methanol (from one of the completed parts of the site) was reduced by only 12 percent. The raids showed, however, that operations over Auschwitz were indeed feasible; only six aircraft were lost despite the strengthening of German defensive measures.¹⁶⁴

The raids against Monowitz took place against the background of a second request for “political” bombing. On August 1, 1944, the Polish Home Army began an armed rising against the German garrison in Warsaw. The Polish army in London requested help from the RAF in the form of military supplies dropped from low altitude over the city. Churchill was once again keen that something should be done.¹⁶⁵ The Operation Frantic shuttle bombing to bases in the Ukraine had been temporarily suspended at Soviet insistence, which ruled out supply missions by the Eighth Air Force. Although Portal and Slessor, Eaker’s second in command in the Mediterranean, regarded the operations as “not practicable” because of the distance and the prospect of high losses, it was decided that the pressure from the Poles and the expectation that the Red Army would soon capture Warsaw were sufficient grounds for undertaking limited operations.¹⁶⁶ The RAF 205 Group, based at Brindisi in southern Italy (considerably closer to Warsaw than bases in England), was ordered to begin nighttime operations. An unofficial mission had already been flown on the night of August 4–5 to drop weapons to Polish partisans, but only six aircraft arrived at the target and four were shot down. On August 8, Moscow was informed that an airlift to the Poles was about to begin, which almost certainly confirmed the Soviet side in the decision not to allow further shuttle bombing until mid-September, when Polish resistance was almost over.¹⁶⁷ On the night of August 8–9 three Polish aircrews successfully reached Warsaw without loss; a total of nineteen missions were flown, the largest on August 14–15 when twelve out of the twenty-seven aircraft dispatched found Warsaw, for the loss of eight aircraft. Total losses were thirty-five bombers (19 percent) out of 195 sorties, but substantial quantities of ammunition and weapons reached the Home Army in the areas of the city where they still held out.¹⁶⁸ In this case the operational conditions were similar to a putative attack on Auschwitz-Birkenau. The difference in the Allied response in the late summer of 1944 can perhaps best be explained in military terms, for the Poles were fighting against the common German enemy. Appeals to help with civilian victims, whether refugees or those slated for genocide, were regarded as outside the remit of Allied military forces, whatever the moral force of the argument. The PWE rejected a Jewish appeal in December 1943 to take action against the Romanians over the killing of Romanian Jews (“considering the constant spate of requests for warning or appeal from Jewish organisations”), but were happy to suggest bombs on Bucharest in March 1944 to speed up the surrender of Romania’s armed forces and to help the approaching Russians.¹⁶⁹ In the end, whether bombing Auschwitz-Birkenau would have had any impact on the conduct of a genocide that had almost run its macabre course by August 1944 remains open to speculation.

The arguments about what was possible in the bombing of eastern Europe, and under what operational conditions, highlights the very different circumstances faced by Allied air forces when confronted with the challenges of distance and geography. For at least the first half of the war, targets in eastern and southeastern Europe were difficult to reach from any bases the RAF might have in the Middle East or North Africa. Navigation problems were magnified for flights from desert airfields across inhospitable terrain with poor mapping and reconnaissance, while maintaining heavy bombers in the heat and dust of the Middle East, thousands of miles from the sophisticated maintenance and logistical system in Britain, was a Sisyphean task. From bases in England, however, most aircraft could not reach distant targets; with the advent of the Lancaster and the Mosquito it took time before serious raids could be mounted even against Berlin, and most of the flight was across the heavily defended areas of the Low Countries and Germany. A large-scale offensive against the Balkan states, Austria, Hungary, and Poland became possible only once bases were available in Italy, from the autumn of 1943.

Some sense of how difficult raiding was to be against targets quite remote from the aerial battlefield in western Europe had already become evident when in 1940, and again in 1941, the RAF undertook preparations to bomb the Soviet oilfields in the Caucasus region in order to deny Germany and Italy vital supplies of fuel. The plans in 1940 were prompted first by the French high command, which wanted to strike at Soviet oil not only to undermine the trade with the Axis states but also to create a possible political crisis for the Soviet Union among the Muslim peoples of southern Russia. French military leaders were much happier about bombing the Soviet Union than bombing Germany.¹⁷⁰ The British side agreed with the plan and drew up a detailed study in April 1940 for deploying forty-eight Blenheim light bombers from bases in Syria and Iraq, supported by sixty-five Glenn Martin bombers bought by the French from American production. RAF planners thought little of Soviet air and antiaircraft defenses, and, like the French, hoped that a three-month attack on Batum, Baku, and Grozny might lead sooner or later “to the complete collapse of the war potential of the USSR,” as well as disastrous repercussions for Germany.¹⁷¹ Chamberlain’s cabinet thought the campaign too risky, and following the German attack on France on May 10, the French abandoned the idea. But the RAF remained in a state of readiness to eliminate the entire Soviet oil industry in three months, assuming an average margin of error of seventy-five yards, a conclusion entirely at odds with all the bombing trials conducted in 1939 and 1940.¹⁷² The plan was revived again in June 1941 in the knowledge that Germany was about to attack the Soviet Union. There were strong recommendations from the British embassy in Cairo and the chiefs of staff to use two squadrons of Wellington bombers and two of Blenheims for a month of intensive attacks, not only to deny the oil to the Germans but “to remind the Soviet of consequences of acceding to German demands.”¹⁷³ Planning was completed by August 1941, but once again operational and strategic reality prevented a campaign in which the means were manifestly inadequate for the military and political ends desired. When an impromptu attack was finally made on German oil supplies in Romania on June 11, 1942, by thirteen B-24 Liberator bombers from the airbase at Fayid in Egypt, the result was described by the Middle East RAF headquarters as a fiasco. The aircraft flew singly and independently; not one reached the oilfield at , but instead they dropped their bombs wherever they could; three returning aircraft landed at Ankara airport, two at Aleppo in Syria, one at Mosul, two more at other desert airfields, and only four reached the planned return base at Habbaniya in Iraq. The unlucky thirteenth aircraft was reported missing.¹⁷⁴

By the summer of 1943 these conditions had altered a great deal. Victory in North Africa in May 1943 opened the way for the invasion and occupation first of Sicily, then of the southern provinces of mainland Italy. Based in Algiers, the Mediterranean high command, first under Eisenhower, then under the British general Henry Maitland Wilson, began to consider at last a full air offensive, combining both political propaganda and bombs, against the Balkan region and more distant targets in central Europe. The North African air forces were transformed into the Mediterranean Allied Air Forces; the American Ninth Air Force (replaced in November by the Fifteenth) was based in southern Italy at Foggia, and the smaller RAF 205 Group at Brindisi. The political offensive mirrored the activity of the PWE and the RAF in western Europe and Germany. It was based on calculations about how populations in the occupied or satellite areas of eastern Europe might react to leaflet propaganda as well as occasional bombing to enhance political pressure. In the Mediterranean, the American Psychological Warfare Branch (PWB) oversaw the production and distribution of most of the Allied political effort in cooperation with officials from the PWE; how that worked in the Italian campaign has already been described. Out of the more than 1.5 billion leaflets produced at the PWB center at Bari and dropped by air or fired in propaganda “shells” whose paper contents burst over enemy lines, hundreds of millions were targeted at Albania, Greece, Hungary, Bulgaria, Romania, Czechoslovakia, and the Yugoslav territories Serbia, Croatia, and Slovenia.¹⁷⁵

The United States demonstrated the same confident enthusiasm for political warfare as the British. “History may well show,” wrote the Eighth Air Force assistant chief of staff, “that no single factor has contributed more to the raising and sustaining of morale in the occupied countries.” The combined effect of the British and American leaflet campaigns, he continued, “will shorten the war as a certainty.”¹⁷⁶ A pamphlet produced under Spaatz’s signature to explain the value of airborne propaganda to American crews (who, like RAF flyers, preferred dropping bombs to paper) claimed that the millions of RAF leaflets had brought “truth, hope and comfort” to the oppressed and sustained the will for sabotage and resistance. “In occupied territory the spirit of rebellion is being fanned,” Spaatz continued. “The output of the factories suffers as surely as if they had been struck by bombs.”¹⁷⁷ A sophisticated technology was developed to ensure that the leaflets fluttered down over a wide area. A single bomber could carry up to a million leaflets at a time. Two large canisters were installed in the bomb bay, each holding sixty bundles of approximately 16,000 leaflets bound by a cord fixed to a barometric device. On release from the aircraft each bundle tumbled down until the change in air pressure acted on the release mechanism, scattering the individual leaflets over a wide area. The system was not foolproof: sometimes the bundles opened prematurely, scattering the loads in the wrong place; sometimes they failed to open and whole bundles, each weighing around fifty-five pounds, fell dangerously on the target population.¹⁷⁸

Both the American PWB and the British PWE understood that for eastern Europe the propaganda had to be carefully calibrated to match the circumstances of individual countries, some of which were satellite states of Germany, others the victims of invasion and occupation. For the satellites the propaganda had to suggest the option of abandoning the German alliance and helping the Allies. The leaflet “Take a Decision” dropped in May 1944 on Hungary, Bulgaria, Romania, and Finland had this object in mind. Occupied Czechoslovakia, on the other hand, had to be appealed to differently. Intelligence sources suggested that the Czechs felt abandoned by the Allies as they had been at Munich in September 1938, and the figures show that Czechoslovakia indeed received only a fraction of the leaflet drops made on other areas.¹⁷⁹ Above all, the political initiative had to be related to the possibility or probability of bombing, either as a threat or as a promise. The Czech sources confirmed that workers were waiting for the bombing to start and would add sabotage to anything the bombs failed to destroy. For the satellite states, bombing was seen, like the case of Bulgaria, as a means to bring the war home to these distant and formerly immune peoples. In the summer of 1944 the Western Allies also had to take account of the onrush of the Red Army, which was by now poised to invade central and southeastern Europe. The PWE assessment of bombing Bucharest, for example, pointed out that bombs might increase Romanian “depression,” but were unlikely to induce defection from the war as the Romanian army struggled desperately to keep the Soviet invaders at bay. Bombs dropped on Hungary were regarded as more useful, as they would remind the Hungarian government and population that they had to do more to sever their connection with Germany.¹⁸⁰ Even against satellite states, the political warfare officers recommended attacks only on evidently military targets so as to avoid alienating the populations that were to be liberated from German domination. Czech informers made it clear that as allies, the Czech people should not be subjected to area bombing, which would provoke “serious resentment.”¹⁸¹ The Yugoslav partisan armies welcomed precise raids on German targets, but not raids on the major cities. The propaganda made much of Allied claims for bombing accuracy against military targets but was occasionally let down by the translation. A leaflet destined for Axis Bulgaria in late 1943 had the English “blockbuster bomb” (designed to destroy factories or military facilities) translated into Bulgarian as “homewrecker.”¹⁸²

These political imperatives were integrated as far as possible with the military planning directed at eastern Europe, although the promise of accuracy was as difficult to fulfill in this case as it had been in the west. For the Western Allies there was only one principal target in eastern Europe once the area came within effective bombing range. The oil-producing region around  in Romania supplied around 3 million tons of oil annually to Germany and Italy out of a total production of 5–6 million tons. For Germany, Romanian exports in 1943 amounted to one-third of all German oil products.¹⁸³ Since oil was a major target for the Combined Bomber Offensive, the interruption of Romanian supplies assumed a high priority. The RAF had begun to explore the possibility of raiding  in the spring of 1942 to aid the Soviet Union, but the operation, at the limit of aircraft range, was regarded as impossible with current strengths. The Combined Chiefs of Staff at the Casablanca Conference in January 1943 called for the immediate bombing of the oilfield, but when Churchill asked Portal to consider the operation he was told that it was still too risky, not only because it would require flying over Turkish airspace to be feasible, but because it would have to be a single heavy and demobilizing strike, which current air strengths in the North African theater could not promise.¹⁸⁴ Although Churchill was willing, as he told Eden, “to put the screw very hard on Turkey” to modify its neutrality for the RAF, the attack on  when it came was made by American air forces acting under pressure from the American Joint Chiefs to act quickly to block Axis oil supplies.¹⁸⁵ The British contribution to the opening raid was to supply good maps of the region and large-scale models of the refineries. Portal was keen to allocate three skilled Lancaster crews because he was not confident that American pilots would be able to navigate the 1,850-mile trip successfully, but in the end the raid launched on August 1, 1943, was made only by the B-24 aircraft of the recently constituted Ninth Air Force.¹⁸⁶

The American operation was first code-named Statesman, then changed in May to “Soapsuds.” Churchill disliked the new choice—“unworthy of those who would face the hazards”—and it was eventually christened, with Roosevelt’s approval, Operation Tidalwave.¹⁸⁷ The operation required a great deal of preparation. It was originally scheduled for June 23, 1943, but postponed not only because priority was given to air support for the invasion of Sicily in July 1943, but because the period of intelligence research and crew training took much longer than anticipated. For the American airmen involved the raid was a daunting prospect. When a British adviser appeared at the airbase at Benghazi from which the raid was to be mounted, he found the morale of the crew “about as bad as it could possibly be”: they told him that they lacked experience of the low-level bombing chosen to maximize the impact of the attack; that they had no previous experience of operations over such long distances; that the countries over which they had to fly were completely exotic, “populated by cannibals” for all they knew. Rigorous training and better information on the value of the raid contributed to overcoming the worst fears, but there could be no disguising that  was one of the most heavily defended targets in Europe.¹⁸⁸ American intelligence on German defenses was in general poor, because the target was remote from the main operational theaters. Instead of the 100 antiaircraft guns identified, there proved to be well over 200; instead of a token force of fighter aircraft, there were more than 200 Me109, Me110, and Ju88 aircraft, as well as the Romanian and Bulgarian air forces along the line of attack. The defense of  was under the command of Lieutenant General Alfred Gerstenberg, who was also the unofficial German “protector” of Romania. He long expected an Allied attack and introduced regular exercises for the defenses as well as establishing a line of radar stations and a corps of visual observers in the Balkan region. Chemical-smoke battalions were ready to obscure the target, while two dummy sites were constructed northwest and east of the complex to distract any attacker.¹⁸⁹ This was as formidable a defense as any available in Germany itself. For the American crews it meant an operation that was likely to be more suicidal than any they would encounter in western Europe.

The raid on August 1, 1943, began early in the morning. Under the command of Brigadier General Uzal Ent, 177 aircraft flew off on a course designed to take them to the northwest of , in order to avoid the guns and the barrage of 100 balloons. Over Romania the lead commander turned east at the wrong point, bringing most of the force close to Bucharest, where the German defenders were put on full alert. The force turned north into the teeth of the antiaircraft and fighter defenses. Some small groups flew low into the oil complex and bombed designated targets from 500 feet, but most, on Ent’s orders, bombed what they could and escaped. The planned return route was abandoned as aircraft damaged by antiaircraft fire and harried by German fighters flew south in disarray. Only 88 returned to Benghazi; 11 landed in Cyprus, 8 in Sicily, 4 on Malta, 8 were interned at Turkish bases, and 2 crashed into the sea. A total of 54 aircraft were lost, many in acts of extraordinary courage in low-level attacks against massed defenses. Almost all those that returned had suffered damage. Two weeks later the survivors were sent on a further long-range mission against the Wiener-Neustadt aircraft plant in Austria, but on this occasion suffered only two losses against the lightest of resistance.¹⁹⁰

The results of the  raid fell short of the ambition to knock the complex out for months, but enough serious damage was done to reduce output substantially at three major refineries and to destroy two completely. Spare capacity and rapid repair nevertheless reduced the effects on German supplies of crude oil. The effect on the local population was to produce a sudden exodus into the surrounding countryside, but casualties were relatively low save for eighty-four women killed when one aircraft crashed onto their prison.¹⁹¹ The oilfield was allowed an eight-month respite during which time the pre-raid levels of output were once again restored. Losses of 40 percent of the force could not be sustained, and the Allied air forces had other urgent priorities in Italy and southern Germany. At a conference on bombing strategy in Gibraltar in November 1943 between Spaatz, Eaker, Tedder, and Doolittle, the main concern was coordinating attacks on German targets from England and Italy. It was agreed that Balkan capitals might make a good morale target, and Sofia was bombed shortly afterward, but oil in southeastern Europe had for the moment disappeared as a priority.¹⁹² Not until March 17, 1944, was Spaatz informed by Arnold that the Combined Chiefs of Staff favored a renewed attack on  when good weather permitted, but the changing strategic situation, with Soviet forces driving on southeastern Europe, gave priority to bombing communications around Bucharest rather than oil, and the target was again postponed.¹⁹³ The change back to oil came later after Spaatz lost his argument with Portal over the best strategy for the pre-Normandy bombing and had to accept the Transportation Plan. In order to get at oil surreptitiously, he allowed Eaker to send aircraft not only against the communications targets around Bucharest but to attack once again the Romanian oilfield.

The result was a devastating series of twenty-four raids against  between April 5 and August 19, 1944, under the command of Major General Nathan Twining. Twenty of the raids were undertaken by the Fifteenth Air Force, four at night by RAF 205 Group. Between them they dropped 13,863 tons of bombs, all but 577 from American aircraft. German and Romanian defenses had been strengthened since the first raid. Alongside thirty-four heavy and sixteen light antiaircraft batteries and seven searchlight batteries, there were between 200 and 250 aircraft.¹⁹⁴ By the close of the offensive there were 278 heavy and 280 light guns, including the new heavy-caliber 105- and 128-millimeter, supported by 1,900 smoke generators. The oilfield was designated a German “stronghold” and Gerstenberg was appointed by the German high command as “German Commandant of the Romanian Oil Region.”¹⁹⁵ But this time the American bombers flew at high altitude, protected by large numbers of long-range P-38 and P-51 fighters. As a result the contest proved more one-sided than the raid in August 1943. Axis air forces managed to mount 182 sorties against the first raid, on April 5, when 13 out of 200 bombers were lost. But by July the sortie rate had fallen to an average of 53 against the five raids that month. On the final raid, on August 19, there was no fighter opposition.¹⁹⁶ Total losses were 230 American bombers, many to antiaircraft fire. Destruction of the refineries was as complete as it could be, with half a million tons of oil destroyed, and more sunk through the successful mining operations on the Danube carried out mainly at night by RAF 205 Group. Some 1,400 mines were dropped and traffic on the Danube was reduced by two-thirds, though 15 percent of the RAF force was lost in raids carried out dangerously at between 100 and 200 feet above the river.¹⁹⁷

In a final gesture, on the eve of the Soviet entry into Romania, Gerstenberg gathered together any German troops he could find, together with the oilfield antiaircraft division, in a bid to seize Bucharest, where the young King Michael had overthrown the Antonescu government, and bring the capital under German control. To support the coup, the German Air Force mounted a heavy raid on the center of Bucharest on August 24, destroying some of the administrative center around the royal palace. The population, according to the German ambassador, was taken completely by surprise, perplexed by the sudden change from German ally to German enemy. But the German coup was stifled by the Romanian army and the German presence replaced by Soviet forces.¹⁹⁸ In September 1944, Eaker was given permission by the authorities in Moscow to visit Romania, which had been occupied by Soviet forces on August 30. He found the devastation at  worse than any photoreconnaissance image he had seen. The information Eaker was given showed that refining capacity had been reduced by 90 percent. By the last attack on August 19, remaining German personnel were only able to transport an estimated 2–4 percent of Romanian capacity. He found his reception cordial and the Red Army commanders astonished at what high-level bombing could achieve. The Romanian people, Eaker reported to Washington, “look upon us as liberators.”¹⁹⁹

In truth the bombing of Romania did not liberate the population but contributed to the collapse of German resistance and half a century of Soviet domination. Whatever the political ambitions to intimidate the Balkan satellites into surrender or to boost the morale of Czechs and Yugoslavs, the pattern of bombing across eastern Europe was to a great extent governed by the military interest of the Allies in weakening German military resistance to the advancing Red Army. The priority given to communications and oil during the course of the summer of 1944 matched the priorities agreed for the bombing of Germany from English bases and brought a great many more locations in Czechoslovakia, Poland, and the Balkans onto the list of key targets assigned to the MAAF. The strategic commitment to attacking German communications in the region was made in April 1944 when it was realized how successful the Soviet advance had been during the winter months.²⁰⁰ The MAAF drew up a survey at the end of April 1944 to see what could be done in “giving direct aid to the Russian army,” first by cutting German supplies, then by interrupting any German withdrawal in the event of a Russian breakthrough.²⁰¹ In May, Portal instructed Spaatz to give top priority to bombing communications in Romania and Hungary and to treat the whole European transport system “as one” when undermining German mobility.²⁰² Eighteen marshaling yards were singled out for attack, with high priority given to the yards at Bucharest where the Romanian authorities told Eaker later that a Fifteenth Air Force raid had killed 12,000 people, 6,000 of them refugees sitting in trains on the track in the belief that the air-raid siren was only a test. The raid was indeed a heavy one, hitting part of the residential area of the city, but official figures showed only 231 killed and 1,567 buildings destroyed or damaged.²⁰³ From June 1944 oil was finally given a top priority and oil targets in Czechoslovakia and Poland were added to the list of potential targets throughout the region.²⁰⁴ Right to the end politics played a part in bombing calculations. Strategic attacks against targets in Germany, Austria, and Hungary could be undertaken either by visual or by blind-bombing techniques against any military target defined in the broadest sense, including “targets of opportunity.” Over Czech and Polish territory crews were only permitted to bomb the designated military target visually or in exceptional cases by blind bombing, but no opportunity targets were permitted, to minimize the risk of casualties among allied populations.²⁰⁵

The most urgent need was to devise a way of ensuring that American and British aircraft did not accidentally bomb or strafe the advancing Soviet lines. In April 1944 a bomb line was agreed, with the Soviet side in southeastern Europe from Constanza, on the Romanian coast, through Bucharest, , and Budapest. Only the last three could be the object of bomb attack, and American airmen were warned to learn the silhouettes and markings of Soviet aircraft.²⁰⁶ The Soviet forces continued to inform their allies of the changing front line through the Allied missions in Moscow, though the MAAF had also posted representatives informally at the headquarters of the Soviet army group in the Balkans to try to minimize any hazard. A zone of forty miles in front of the advancing Red Army was agreed as the limit for British and American bomber operations, but information about where the line was had to come through a cumbersome process of consultation in the Soviet capital. As the Allied air forces converged from east and west, the danger of inadvertent bombing increased. On November 7, 1944, a force of twenty-seven P-38 Lightnings strafed and bombed a Soviet column in Yugoslavia fifty miles behind the Soviet front line, killing the commander and five others. Three of the nine Soviet fighters sent to protect the column were shot down.²⁰⁷ Stalin’s military headquarters made a strong protest and suggested a bomb line running from Stettin on the Baltic coast down through Vienna to Zagreb and Sarajevo in Yugoslavia, leaving many designated oil and transport targets out of bounds to Western air forces. The Combined Chiefs of Staff refused to accept a bomb line farther north than the Danube, but offered to set up a proper liaison organization with advancing Soviet armies to avoid further mishaps. When Moscow rejected the idea of any collaboration on the ground, Spaatz and Eaker worked out their own bomb line and communicated it to Moscow, sealing Dresden’s fate a few weeks later.²⁰⁸ The Soviet desire to reduce the bombing in eastern Europe was not disinterested, since by that stage of the war Moscow wanted to capture resources, equipment, and plants intact before the strategic air forces obliterated them shortly before they fell into the Soviet sphere. After the war, the formal communist line was to argue that bombing in the teeth of the Red Army advance had been carried out by the agents of capitalist imperialism to weaken the future socialist economy.

Unlike the situation in the western zones of Germany and Austria, it proved impossible for American or British intelligence teams to survey systematically the damage that bombing had done to the industrial and infrastructure targets in eastern Europe, or to establish how effective the political offensive against the satellite and occupied populations had been. An American military mission arrived in Sofia in November 1944, but the Red Army command proved uncooperative. Eaker’s visit to  in September 1944 was the closest that Allied air commanders got to assessment of the damage, and his judgment that the offensive was “a perfect example of what bombing can do to industry” is supported by the German figures on oil supply.²⁰⁹ By the end of the war relations between the Allies were already cooling and Stalin was unwilling to allow Western intelligence officers access to the bomb sites in the Soviet-controlled regions of Europe. In July 1945 some of the USSBS team arrived in Berlin, where amid the chaos they tracked down Speer’s chief economist, Rolf Wagenführ, who was already working for the Russian occupiers. An American team broke into his house in Soviet East Berlin, dragged him out of bed, and bundled him onto an airplane to the American zone, where he gave advice on German statistics for two weeks before being sent back. A key was also found to the German Air Ministry document safe where additional information was discovered; a discreet foray into the Soviet zone secured more German papers.²¹⁰ But all this was little substitute for ground-level reconnaissance of the targets bombed in Czechoslovakia, Poland, and the Balkan states. Assessment of whether bombing had delivered the political dividend in encouraging Axis populations to abandon the link with Germany was open to speculation. In Bulgaria, Romania, and Slovakia the political scene was dominated by the imminent arrival of the Red Army. The resentment and anxiety provoked by sporadic bombing of civilian areas, evident from intelligence sources, paled into insignificance at the prospect of a Soviet empire.²¹¹ This was not the political outcome the West had wished for when the air forces were at last in the position to rain down bombs and leaflets on the distant East.

Rotterdam Once Again

The situation for the smaller states in the German New Order on the northern fringes of Europe—Belgium, the Netherlands, Denmark, and Norway—differed from the fate of France and of eastern Europe. In both these latter cases invasion displaced the German occupier well before the final end of the war in Europe. Belgium was finally fully liberated by November 1944, but not before heavy bombing by the Allies and V-weapon attack by the German side had inflicted wide destruction and casualties. The Netherlands, Denmark, and Norway remained under German occupation until the end of the war. A raid by RAF bombers that hit residential districts of The Hague on March 3, 1945, killing more than 500 people, occurred just weeks before liberation.²¹² Over the course of the war the enthusiasm relayed to the Allies by Dutch, Belgian, or Norwegian resisters about bombing German targets became tempered by growing resentment at the cost in lives and livelihoods imposed on those caught in the crossfire of war.

There was also an important political difference in the case of Norway, the Netherlands, and Belgium. Each had a government-in-exile in London, with a miniature apparatus of state. Unlike any other bombed state, the exile governments could represent directly to the British government their views on bombing policy and their objection to or approval of its conduct. This in turn placed considerable pressure on the RAF, and later the U.S. air forces, to ensure that the guidelines governing the bombing of targets in the region clearly expressed its limits. Damage to civilian targets and civilian deaths had to be explained or apologized for, unlike raiding against most other targets in Europe. This was even more the case when bombs intended for a German destination instead fell inadvertently on Dutch or Belgian cities. Both states lay on the flight paths to German targets for the whole war period. The Dutch town of Groningen was bombed by mistake on July 26, 1940, and two people were killed; it was bombed again in error when the German port of Emden was shrouded in fog on the night of September 26–27, 1941, and this time six people died.²¹³Maastricht was bombed by mistake for Aachen in February 1942, prompting Dutch protests that the RAF used trainee crews for nearer and easier targets, as they did, making mischance more likely. Raids deliberately targeted against Dutch cities provoked even higher casualties. Operations against targets in Rotterdam in October 1941 and January 1942 cost 177 Dutch lives; a raid in October 1942 on Geelen and one two months later on the Philips electrical works at Eindhoven killed a further 221.²¹⁴ Although the British Foreign Office believed that the Dutch took the robust view that “war is war,” Eden remained keen to ensure that proper guidelines were drawn up and crews instructed in their observance.²¹⁵ Belgian and Dutch targets were governed by the same rules drawn up in October 1942 for bombing occupied France, with the difference that attacks on trains by night could only take place between the hours of eleven o’clock in the evening and four o’clock in the morning rather than throughout the hours of darkness. Military targets had to be identified, and if the prospect of a “large error” occurred that was likely to lead to civilian casualties, the operation was supposed to be aborted.²¹⁶

The systematic bombing of military and industrial targets in the Low Countries only began in 1943 when the Eighth Air Force used its heavy and medium bombers for attacks on targets near enough for fighter protection so that novice crews could be initiated into operational practice with fewer immediate risks, the very policy condemned by the Dutch when bombing started in 1941. The result was an immediate disaster. A daylight raid on Rotterdam on March 31, 1943, cost an estimated 400 lives; an attack that was supposed to hit the German Erla aircraft plant near Antwerp instead devastated the town of Mortsel, killing 926 Belgians, including 209 children in four schools hit by the bombs. The bombing of Mortsel resulted in the worst casualties of the war from a single raid on the Low Countries. No warning leaflets had been dropped and the town was crowded with people. In addition to the dead, a further 1,342 were injured, 587 seriously, while 3,424 houses were damaged or destroyed. In the Erla works, 222 workers were killed.²¹⁷ Although heavy damage was done to the aircraft plant, the bombs were strewn over a wide residential area. The Eighth Air Force post-raid assessment showed that only 78 out of 383 bombs dropped came within 2,000 feet (600 meters) of the target.²¹⁸ For the residents of the town the raid was a shocking realization of the horrors of aerial war. “I heard the screams of dying schoolchildren,” recalled one witness. “I heard the grief-stricken cries of desperate mothers and fathers, searching in the ruins for their beloved children. . . . I saw fires, heaps of ruins and people wringing their hands.”²¹⁹ A Foreign Office official wrote to the PWE a few days later describing the raid as “catastrophic”; the “bad shooting” of the Eighth Air Force, he complained, had done serious damage to the reputation of Allied air forces in a country that had hitherto welcomed the idea of bombs on German targets.²²⁰

The poor record of American bombing soon provoked a wider crisis. Eaker was first asked by the Air Ministry to take special care in bombing centers where civilians were likely to be hurt, then told to suspend bombing in occupied areas altogether pending a decision about which specific targets should be allowed. Despite protests from Washington, a list was agreed to between the two air forces, and Eaker and Harris were asked to avoid using freshmen crews against targets in major cities. The twenty accepted targets included only one in the Netherlands and five in Belgium. Bomber Command agreed to use only reliable and experienced crews, but Eaker insisted on the continued right of the American bomber force to use novices.²²¹ No sooner was the list agreed upon than the new Pointblank directive, drawn up for the Combined Bomber Offensive, apparently reversed the decision by listing targets in densely populated areas of occupied Europe as part of the overall plan. Eaker at once asked whether he was now free to bomb what he liked, but the issue could only be resolved at the highest level. Eden worried that civilian casualties in Belgium and the Netherlands would deeply affect “the morale and spirit” of the local population and asked that radio broadcasts and leaflets should notify the people of impending bombing, but only after agreement had been reached with the Dutch and Belgian governments in London.²²² Leaflets were drafted warning the population that it would be dangerous to work in any factories assembling aircraft, locomotives, submarines, and vehicles or any one of their component parts. On June 25 the Belgian government-in-exile agreed to allow the new targets to be bombed once warning had been given; the Dutch government followed suit on July 15, but only after making it clear that they would only tolerate operations conducted in such a way “as to minimize the danger to the civilian population.”²²³

Once again the raiding habit of flying at high altitude against targets that even when visible could not be hit with sufficient accuracy broke the pledge to bomb with greater discretion. The Eighth Air Force began operations with B-17 Flying Fortresses against Dutch targets just two days after receiving approval from the Dutch government. The target chosen was the Fokker aircraft plant in Amsterdam, first bombed by the RAF in 1940 with little effect.²²⁴ The American operation on July 17 killed 185 people and missed the factory. One bomb hit the church of St. Rita, filled with 500 schoolchildren who were singing an “Ave Maria” to ward off danger after the siren had sounded. Eleven were killed in the church; another 29 died when a bomb hit a doctor’s waiting room. Around 130 buildings were destroyed.²²⁵ The Dutch government-in-exile immediately protested and Eaker was asked to explain how he was going to avoid a repetition. The Eighth Air Force switched temporarily from using B-17s to using the medium twin-engine Martin Marauder B-26 bombers for attacks on Dutch and Belgian targets from lower altitude. On their second raid, against a Dutch power station at Ijmuiden, all eleven B-26s were shot down.²²⁶ Over the months that followed the B-26s were instructed to fly higher and casualties on the ground mounted again. The bombing of Ghent on September 4, 1943, resulted in 111 dead, that on Brussels on September 7 using B-17s caused a further 327 deaths.²²⁷ An American raid on Enschede on October 10, again using B-17s, killed 150. The damage to the German war effort was limited. The Fokker works completed a program of dispersal and decentralization into forty-three smaller locations scattered around the Amsterdam region.²²⁸ For Dutch workers and producers, as in Belgium and France, the choice of refusing to work for the German military was to run the risk that both the machinery and the workers would be transferred to Germany. In Rotterdam by 1944 over 40,000 workers had been sent to work in the Reich; repeated raids, which killed a total of 748 people in the city during the war, encouraged the German occupiers to move workers to industry in Germany, where there was effective protection and the means to compel compliance.²²⁹

The inclusion of Dutch and Belgian firms in the Pointblank plan came about as a result of the contribution made by their industries to German aircraft and submarine production, as well as the supply of machinery and steel. There were no complete aircraft produced in Belgium, although hundreds of small firms supplied components; in the Netherlands, however, 414 aircraft were built in 1943 and 442 in 1944, while Dutch shipbuilders supplied an important source of additional capacity for the submarine industry and for the production of smaller naval vessels.²³⁰ By the end of 1943 around 75,000 Belgian and 109,000 Dutch workers were employed on German arms contracts.²³¹ The transfer of German production to the occupied territories gave the occupiers sufficient reason to provide limited protection from air attack. Antiaircraft guns and fighters stationed in the Low Countries formed, as in France, part of the air defense rampart around the European fortress, which in the Netherlands included ten squadrons of night fighters by 1944 and seventy-four sites for radar and electronic warfare.²³² Regiments of German fire-protection police were also stationed in the Low Countries to supplement the efforts of the local population organized, in the Dutch case, in an Air Protection Service (Luchtbeschermingsdienst) first set up in April 1936 and organized by urban district, street, and housing block. The attitude of the population to the raiding was complex, since many Allied aircraft crashed in the Low Countries, and in many cases surviving crew were helped by the local population or benefited from escape networks. The local press generally condemned the raids with high civilian casualties, although in no occupied territory was the press free from German controls or from the numerous German communiqués on terror bombing they were given. An article in the Haagsche Courant on a bombing attack with twenty-two dead carried the headline “That’s It: Murder,” and ended with the question “Is that not terroristic?” Another article following the bombing of Rotterdam on January 29, 1942, carried the headline “Bloody Work of the English Air Pirates!”²³³ British intelligence from Belgium and the Netherlands, however, only suggested that the barometer of popular support for bombing fluctuated as it did in France with the perceived accuracy or inaccuracy of a raid. There was, however, little sign of open resistance as there had been in the early years of war. A report sent to the PWE in November 1943 observed that the people “are just weak and passive.”²³⁴

The most intensive period of bombing came in 1944 with the preparations for the Normandy invasion and the Crossbow raids against sites intended for the V-1 and V-2 missiles. Most of the raids were small tactical raids against rail and air targets carried out by medium bombers and fighter-bombers. But as Belgium and the Netherlands became involuntary parts of the vast aerial battlefield in northwestern Europe, so the pace of destruction and loss of life quickened. There descended on both populations a rain of bombs and leaflets, the second designed to explain the liberating effects of the first. In April and May 1944 the Eighth Air Force undertook 1,111 sorties against German airbases in the two countries, and in May 759 sorties against marshaling yards, losing only twenty aircraft in the process.²³⁵ The preparatory raids by heavy bombers involved some of the highest casualties of the war. In Belgium 252 were killed in Kortrijk on March 26, 428 in Ghent on April 10. The peak in Belgium came between May 10 and 12, when more than 1,500 died, culminating in a raid on Leuven in which 246 were killed. The Overlord preparations cost Belgium a total of 2,180 civilian dead.²³⁶ The Netherlands was bombed less heavily during the late spring, but on February 22, 1944, it had suffered the severest raid of the war when the city center of Nijmegen was hit by a group of B-24 bombers returning from an aborted mission over Germany. The aircrew aimed for a marshaling yard on the edge of the town, still believing that they were over German territory. Instead the lead bombardier misjudged the speed, dropping bombs in the crowded city center, followed by the rest of the combat box. The estimated 800 dead were caught in the open after the all-clear had been given a few minutes before the bombs began to fall. Some 1,270 buildings were destroyed or badly damaged. On this occasion the Dutch government, not fully aware of the results of the raid, lodged no protest.²³⁷ Over the following three months more than 50 million leaflets were dropped on Belgium, 55 million on the Netherlands, preparing both populations for their liberation but including a leaflet apologizing for the bombing of Nijmegen and regretting that under the circumstances of modern air war “sometimes harm and grief was caused to our friends.”²³⁸

During the period when the noose was tightened around the German New Order, little attention was given to conditions in Scandinavia. Denmark, like the Low Countries, was astride the bombing runs to targets in Germany and became the inadvertent target from occasional errors of navigation as well as the site of numerous Allied aircraft that either crashed or were forced to land. Most of the raiding against German naval targets and airfields was carried out by aircraft of Coastal Command until the end of 1941, when Bomber Command Blenheims and Mosquitos were detailed to attack land targets, while Coastal Command concentrated on targets at sea. The German occupiers collaborated with the Danish civil defense organization, Statens Civile Luftvaern, in observing incoming aircraft. A large number of light antiaircraft guns protected military installations, and when the Kammhuber Line was extended into Denmark to give better protection against RAF raids in 1941, squadrons of German night fighters were also based at Danish airfields.²³⁹ There were almost no strategic Allied air attacks on Denmark. Most bombs were jettisoned or dropped in error, a total of 3,269 high explosive and 22,298 incendiaries counted by Danish civil defense. The only planned attack on an industrial target took place on January 27, 1943, when eight Mosquitos attacked the Burmeister & Wain diesel engine factory in Copenhagen. Some damage was done, and a sugar plant was set on fire. The raid resulted from PWE pressure to undertake an attack on at least one industrial target in each occupied state to discourage collaboration and to reinforce local morale. In the Danish case the raid proved a success: there were few casualties, and in the aftermath Danes took to wearing RAF colors as a mark of sympathy with a distant ally. The only raid by the Eighth Air Force was against Esbjerg airfield on the west coast of Jutland on August 27, 1944, bombed as a target of opportunity following an aborted raid on Berlin. Over the whole course of the war 307 Danes were killed as a result of all Allied air activity and 788 injured.²⁴⁰

Operations against Norway were also linked closely to the air-sea war, but Norway differed from Denmark because there were important industrial and raw material resources that the Germans exploited throughout the occupation, particularly aluminum production and the development of “heavy water” for the German nuclear research program. Both were attacked by the Eighth Air Force, the aluminum plant at Heröya on July 24, 1943, the Norsk Hydro plant on November 16. Spaatz was pleased with the result and thought “it was heartening to the Norwegians.”²⁴¹ Far from being heartened, the Norwegian government complained that the destruction of the Norsk Hydro plant, which produced a large quantity of Scandinavia’s much-needed fertilizer, had provoked “bewilderment and dismay” among the Norwegian population as they contemplated declining food output. The small heavy water plant had been successfully sabotaged by the Norwegian Resistance some months before. The Norwegian foreign minister, Trygve Lie, asked the Allies to agree upon a means of collaborating on the choice of targets to avoid further mishaps. It took almost a year of argument before an agreed list was drawn up. British air marshals, Lie complained, were a law unto themselves.²⁴² When the Air Ministry presented a list of seven targets they would like to bomb, the Norwegian high command responded that some were wrong, some had ceased operating, and others were essential to Norway’s economy when the war was over.²⁴³ A second list was worked out with Norwegian advice and finally agreed to on November 2, 1944. But only four days before that, Bomber Command had tried to hit the submarine pens at Bergen in poor weather. The bombs struck the town center, killing fifty-two civilians and burning down Europe’s oldest theater. The Norwegian government again warned the Foreign Office that raids without an evident military purpose merely alienated a potentially friendly population. Though on the agreed list, Bergen had been attacked by forty-seven Lancasters, against instructions, through almost complete cloud cover.²⁴⁴

For Norway and Denmark the price of remaining occupied was substantially lower than the price paid by Belgium and the Netherlands as they became the focus of the ground campaign from the autumn of 1944. As was the case in France, operational requirements soon came to replace political calculations when deciding on targets to attack. Belgium was caught in between the fighting powers as Allied armies occupied Belgian territory in September 1944. From sites in western Germany, the V-weapons were turned against Allied forces in Antwerp and the surrounding territory, while a few were launched against Paris. The first V-2 rocket fell on October 7 in Brasschaat on the outskirts of Antwerp, the first V-1 flying bomb on October 21. The last V-1 struck on March 28, 1945. Around 12,000 V-1s were launched at Belgian targets, and 1,600 V-2s.²⁴⁵ The port of Antwerp suffered most. The worst incident occurred on December 16 when a V-2 fell on the Rex Cinema, killing 271 Belgians and an estimated 300 soldiers.²⁴⁶ Late in 1944 a British civil defense mobile column was sent to Belgium at the request of Montgomery’s 21st Army Group, complete with canteens, ambulances, and fire units; three British rescue instructors were sent to Brussels, Antwerp, and Eindhoven to train soldiers for emergency work after a rocket strike.²⁴⁷ In November 1944 the American brigadier general C. H. Armstrong was appointed commander of Flying Bomb Command Antwerp X, and three belts of antiaircraft guns were set up in a ring around the city to shoot down the flying bombs. RAF Fighter Command set up a Continental Crossbow Forward Unit in December 1944 to add fighter interception of the flying bombs to the effects of antiaircraft fire. By February almost three-quarters of the V-1s were destroyed before they hit the city, a total of 7,412 over the course of the campaign. Only 73 fell in the dock area of Antwerp and only 101 on the built-up area. The effect on the flow of Allied supplies through the port was described in the official account of the campaign as “negligible.”²⁴⁸ Against the V-2 rocket, however, there was no defense; the Belgian population became once again hostage to their geographical location and suffered a heavy toll. In total 6,500 were killed and 22,500 injured in the last flurry of bombardment from the air, almost exactly the same number of casualties exacted by the V-weapon attacks on England. This final German aerial assault greatly increased the overall Belgian casualties from air bombardment during the war. An estimated total of 18,000 Belgians were killed, one-third from German operations.²⁴⁹

Since the Netherlands was used as a base for firing V-weapons, the threat of Allied bombing hung over the Dutch population until almost the end of the war. The battle for Arnhem in September 1944 brought further heavy raids against German military targets by both the Eighth Air Force and Bomber Command. The continual bombardment of London by V-2 rockets from sites in the Netherlands finally prompted a decision to try to neutralize the threat by bombing, despite their proximity to residential areas. On March 3, RAF bombers of the Second Tactical Air Force flew from bases in Belgium to bomb a V-2 site in a large park in the north of The Hague. The weather was poor—cloudy and with a strong wind—and the briefing officer had confused the map coordinates, instructing the force to bomb more than a kilometer away from the intended target. The sixty B-25 Mitchell and A-20 Boston/Havoc aircraft dropped sixty-seven tons of bombs on a residential area of the Dutch capital, killing an estimated 520 and rendering 12,000 homeless. A stream of up to 50,000 refugees fled from the area, some of them, one eyewitness wrote, still in their pajamas: “a long procession of people and children crying . . . some were all white from lying under the ruins. Others bled from a variety of injuries, which were half-bandaged or not covered at all. . . . More and yet more.”²⁵⁰ Over the course of the war an estimated total of between 8,000 and 10,000 Dutch deaths came from bombing, around one-tenth of them caused by German raids.²⁵¹

The bombing, so close to the end of the war, provoked a furious response from the Dutch government in London and strong criticism from Churchill. A broadcast apology was made later in the month, promising a full inquiry, but nothing was relayed to the Dutch authorities, who repeated the request in June 1945 for an explanation.²⁵² The Air Ministry told the Foreign Office at the end of June that the internal investigation had discovered the error in plotting the coordinates for the raid and had court-martialed the officer responsible. There was, however, little sense of contrition. The operation, claimed the ministry, was a difficult one: “The extent of the disaster must to some extent be set down to the mischances of war.”²⁵³ Six months later a Dutch woman wrote to King George VI asking him to pay compensation for the total loss of her house and possessions in the March 3 raid. The letter summed up the ambivalence felt by the liberated peoples about being bombed into freedom:

I humbly come to you, first to express my great gratitude for all you, the English Government and the English people have done to deliver us from those awful huns. Second to ask you for help. On March 3 my house (home) with all that was therein, was bombed and nothing could be saved. . . . And now after nearly ten months, I sit here as poor and forlorn as on March 3. . . . It may seem rather impudent from me, to ask you for help, but I know you are righteous and honest, above all, and that in no way you will have a widow been left in solicitude [sic] and affliction, where there is still a debt for the RAF to be redeemed.²⁵⁴

The Foreign Office contacted the Air Ministry for confirmation that nothing could be done for what was clearly “a hazard of war.” The ministry replied that nothing should be done: “If we started paying for this kind of loss there would be no end to our liability.”²⁵⁵

•   •   •

In all the bombing of New Order Europe a balance was supposed to be struck between political calculation and military imperatives, since the peoples to be bombed were allies or potential allies. The cautious bombing of the first two years of war reflected a balance in favor of political restraint. In some cases bombing was used as a “calling card,” to remind populations under German domination not to collaborate or to encourage confidence in eventual liberation. In some cases the resistance movements called for bombing because they accepted the airmen’s claims about its accuracy and power. However, by the spring of 1942 and on to the end of the war, the pendulum swung slowly in favor of military necessity, while European resisters became disillusioned about what bombing might achieve. In June 1943, Sinclair asked Eden to reconsider the principle that causing casualties to non-German civilians was a sufficient ground for restraint. Eden had a politician’s instinct that killing Allied civilians was the wrong course, but his reply to Sinclair symbolized the shift in priorities: “If the new bombing plan is strategically necessary, I shall not of course stand in its way.”²⁵⁶ American airmen were in general less affected by political calculation, partly because the State Department was geographically remote, partly because Americans were outsiders in the European theater, less aware of the political realities they faced. As the western war entered Europe, bombing became more widespread and its effects usually indiscriminate. By 1944, Allied commanders were increasingly “bomb happy,” summoning bombing whenever there was a problem to solve. This produced dangerous paradoxes for the peoples of the New Order: the closer to victory and liberation, the more deadly became the bombing; as the bombing intensified, so German antiaircraft resources were spread ever more thinly around the perimeter of the German empire, exposing the subject peoples more fully to the rigors of bombardment. The occupied states had their own civil defense organizations, but they were in general less well resourced than in Britain or Germany. This would have mattered less if the bombing of military objectives had been as accurate as the Allies claimed (and on occasion achieved). The bombing of Brest, Le Havre, Caen, Mortsel, The Hague, Bucharest, and a dozen or more other cities exposed the hollowness of any claim to operational precision. Bombing was a blunt instrument, as the Allies knew full well, but its bluntness was more evident and more awkward when the bombs fell outside Germany.