by Paul Kennedy
LOCATION OF MERCHANT SHIPS OF THE BRITISH EMPIRE, NOVEMBER 1937
A wonderful scattergraph map showing the sheer enormity and vulnerability of the British Empire’s merchant shipping across the entire globe. Note the significance of the Mediterranean route, the massive trades from the Caribbean and South America, and the strategic significance of Freetown, Sierra Leone, and Gibraltar. This free-ranging system could not sustain the ravages of constant U-boat and aerial attack.
Click here to download a PDF of this map.
In addition, although the Admiralty did not possess an exact tally of the enemy’s U-boats, there did seem to be an awful lot more of them. In the course of 1942, Allied warships and aircraft had destroyed eighty-seven German and twenty-two Italian submarines. But the Third Reich was also gearing up its war production and had added seventeen new U-boats each month during that year. By the end of 1942, therefore, Doenitz commanded a total of 212 submarines that were operational (out of a grand total of 393, for many were working up, training new crews, or receiving new equipment), very significantly more than the 91 operational craft he had had (out of 249) at the beginning of that year.2 Although victory in the Second World War was critically affected by each side’s inventiveness, technology, and organization, not just by sheer numbers, the blunt fact was that numbers did count. And by the time of the Casablanca conference it seemed that the Germans were having greater success at sinking Allied merchantmen than the Anglo-American forces were in sinking U-boats. Worse still, more and more U-boats were entering the fray.
In the months that followed, therefore, the prime minister’s nightmare appeared to be coming true. As March and April 1943 arrived and the convoy traffic to the British Isles resumed at a higher rate, so too did merchant ship losses. February’s total doubled that of the previous month, and in March the Allies lost 108 ships totaling 627,000 tons, making it the third-worst month on record during the war. What was more, nearly two-thirds of those ships were sunk in convoy; one was no longer talking here of the happy U-boat pickings of individual merchant ships off the well-lit shores of America early in 1942, or of the almost equally easy raids upon Allied shipping routes in the South Atlantic. What was also truly alarming was that the losses had occurred chiefly along the single most important convoy route of all, that between New York and Halifax and the receiving ports of Glasgow and Liverpool. Between March 16 and 20 the greatest encounter in the entire Battle of the Atlantic saw Doenitz throw no fewer than forty U-boats against the two eastbound convoys HX 229 and SC 122. This epic fight will be analyzed in more detail below, but the result was awful for the Allies: twenty-one merchant ships totaling 141,000 tons were sunk, with the loss of only one U-boat. In the Admiralty’s own later account, “The Germans never came so near to disrupting communications between the New World and the Old as in the first twenty days of March, 1943.”3 These ever-rising losses suggested, among other things, that the whole principle of convoy as the best means of defending maritime commerce was now in doubt.
The Strategic and Operational Context
The British Admiralty’s problems were nothing new in the annals of naval warfare. The protection of ships carrying goods at sea from hostile attack is one of the oldest problems in the history of war and peace. Even at the height of the Roman Empire merchants and consuls in Sicily and North Africa complained about the depredations of pirates against the grain, wine, and olive oil trades. Fifteen hundred years later Spanish commanders fumed at the plundering by Dutch and English raiders of their galleons bearing silver and precious spices; only a generation or two afterward the Dutch found their long-haul trade with the East Indies under French and English assault at sea. The age of European expansion and then of the Commercial Revolution (sixteenth to eighteenth centuries) had moved ever greater shares of national wealth onto precarious maritime routes. In the age of Charlemagne, the dependence of rulers and peoples upon command of the sea was negligible. By the time of, say, the Seven Years’ War (1756–63) it was critical in both the Atlantic and Indian Oceans, at least for all advanced economies. If a west European nation lost control of the trade routes, it was most likely also going to lose—or at least not win—the war itself. This was the message of that classic work, The Influence of Sea Power upon History (1890), composed by the American naval author Alfred Thayer Mahan.
Mahan’s ideas influenced the admiralties of Britain, Germany, Japan, the United States, and many lesser navies. The key belief was that the only way to gain command of the sea was to have the most powerful battle fleet afloat, one that would crush all rivals. Lesser forms of naval warfare, such as commerce raiding and cruiser and torpedo-boat operations—la guerre de course—didn’t count for much, because they didn’t win wars. It was true that during the Napoleonic Wars French predators had seized many independently sailing British merchantmen, but once the latter were organized in convoys and given an escort of warships, the sea routes were secure behind Nelson’s assembled fleets. The same truth revealed itself, albeit at great cost, during the First World War. For three years, and even though the Grand Fleet had command of the sea, Allied merchant ships steaming on their own were picked off in increasing numbers by German U-boats. After the Admiralty was compelled by the British cabinet to return to the convoy system in 1917, losses to enemy submarines dropped dramatically. Within a short while, moreover, the Allied warships would possess asdic (sonar), so for the first time ever they could detect a solid object under water. Provided one had command of the sea on the surface, it was argued, one would also control the waters below. A submarine would thus be as recognizable as the sails of a French frigate 150 years earlier. Such was the prevailing assumption of naval staffs in the years following the Treaty of Versailles, 1919. Convoys, plus sonar, worked.4
Before we examine how and why that assumption was challenged by the renewed German U-boat threat during the first half of the Second World War, a couple of very important, though clashing, strategic-operational assumptions also need to be considered. The first of these, rarely articulated, is that one really didn’t need to sink surface commerce raiders or submarines to win the maritime war. So long as the Royal Navy shepherded without loss a group of fifty merchantmen from, say, Halifax to Liverpool, it had won. The larger Allied strategy was to keep Britain in the fight and then to make it the springboard for an enormous invasion of western Europe. Thus, if every transatlantic (and South American, Sierra Leonean, and South African) convoy got to port safely without ever encountering U-boats, the war was being won, ship by ship, cargo by cargo. Even if the convoy escorts had to face a serious submarine attack but could beat off the predators, all would still be well. The task of the shepherd was to safeguard the sheep, not to kill the wolves.
The opposite argument was that killing the wolves had to be the essence of Allied maritime strategy. It too had its own logic: if the threat to the sea-lanes was forcibly removed, all would be fine and one of the Casablanca war aims could at last be implemented. In today’s language, the prevailing authorities cannot wait for terrorists to attack the international system but have to go and root out the terrorists. In maritime terms, therefore, a navy charged with protecting its merchant ships would either go on a submarine hunt or, an even bolder tactic, simply drive its convoys through U-boat-infested waters and force the submarines to fight—and be killed.
The first of these two convoy strategies was clearly defensive; the second (whether submarine hunting or forcing the convoys through) was equally clearly offensive. Both visions, it is worth noting, involved a tricky, interdependent three-way relationship between the merchant ships, the U-boats, and the naval and aerial escorts, not unlike the children’s rock-paper-scissors game. If the convoys could avoid an encounter or have the U-boats beaten off, fine for them; if the U-boats could get at the convoys without destruction from the escorts, fine for them; and if the escorts could destroy enough submarines, fine for them.
In the harsh world of the North Atlantic between 1939 and 1943, however, neit
her an Allied defensive operational strategy nor an offensive one was possible on its own. The way forward had to be achieved by a combination of both options, depending on the ups and downs of what turned out to be the longest campaign of the entire Second World War. And this route was, geopolitically, the most important maritime journey in the world. Of course the other Allied trade routes mattered, and all faced the same operational and logistical difficulties, or sometimes (as with the Arctic convoys) even greater ones. But maritime security across the Atlantic was the foundation stone of all Anglo-American grand strategy in the European theater. With a look forward to the remaining chapters of this book, it is worth restating the many interconnections. Winning this Atlantic battle preserved Britain’s own very large military-industrial base. Britain was also the unsinkable aircraft carrier for the Allied strategic bombing campaigns, and the springboard for the eventual invasion of western Europe. Britain was the port of departure for most of the convoys to northern Russia and to the Mediterranean; it was the source for the many troopship convoys that Churchill dispatched, via the Cape, to Montgomery in Egypt and the Middle East. Controlling the Atlantic was the sheet anchor of the West’s plans to defeat Italy and Germany.
The top-to-bottom logic chain of Allied grand strategy here is also unusually clear, a fine example of Millett and Murray’s concept of the multilevel nature of “military effectiveness.”5 The political aim was the unconditional defeat of the enemy, and the return to a world of peace and order. The strategy to achieve that purpose was to take the war to the enemy by all the means that were available: aerial, land-based, naval, economic, and diplomatic. This required decisive successes at the operational level, and in all the areas covered in the chapters of the present book. It would be foolish to argue about which of those operational regions was more important than others (even if the Combined Chiefs of Staff had to do so as they wrangled over allocating resources); they all were part of Allied grand strategy. What is incontestable, however, is that if the British, the Americans, and their smaller allies were to reconquer Europe from fascism, they first of all had to have command of the Atlantic waters.
Yet control of that vital route was itself determined, in the last resort, by a number of key technical and tactical factors. In other words, there is also in this story a clear example of a bottom-to-top logic chain. Every individual merchant ship that was preserved and every individual U-boat that was sunk by Allied escorts directly contributed to the relative success rate of each convoy. The tactical success rates of each single convoy contributed to the all-important monthly tonnage totals, and those monthly tonnage totals were the barometer to the winning or losing of the Battle of the Atlantic. That operational battle, as we have argued, was key to victory in western Europe and the Mediterranean. And winning in the West was a part of the strategic tripos—victory in the West, victory in the East (Eastern Front), and victory in the Pacific/Far East.
The Battle of the Atlantic was an operational and tactical contest that hung upon many factors. The first of these, from which all the others flowed, was the possession of efficient and authoritative organization. This was so basic a point that it is often taken for granted, yet on brief reflection it is clear how important were the structures of command, the lines of information, and the integration of war-fighting systems. Both sides benefited greatly, of course, from the experiences of the epic campaign in the Atlantic during the First World War, and by the post-1919 lessons drawn from them. In terms of simplicity of command, Doenitz had it easier, for the U-boat service was separate from the German surface navy, and it became easier still for him when the failure of a squadron of heavy ships to destroy an Arctic convoy in the last days of 1942 led to an explosion of rage on Hitler’s part and to Grand Admiral Erich Raeder being replaced as commander in chief of the entire navy by Doenitz himself at the end of January 1943. Doenitz decided to remain commander of U-boats, so as to keep control of submarine operations, and it is evident that he found it much easier to obtain the Fuehrer’s backing than his predecessor had. This did not mean that he had no organizational fights. There was a constant struggle to gain the necessary share of war materials (steel, ball bearings, electrical parts, antiaircraft weaponry) against the enormous demands of the Wehrmacht and the Luftwaffe. And, as we shall see, Doenitz had the greatest difficulty in getting aerial support for his boats. Nonetheless, it was an enormous advantage to have a single and very experienced authority directing the entire U-boat campaign.
The organizational story on the Allied side was a lot less clear-cut. During the years of American neutrality, it actually had been simpler. Ensuring command of the seas was the traditional responsibility of the Admiralty in Whitehall, which then devolved defense of the Atlantic convoys to a particular authority, the Western Approaches Command, based in Liverpool. By the time our analysis begins, its chief was the formidable Admiral Sir Max Horton, like Doenitz a highly experienced submarine commander twenty-five years previously. The much smaller Royal Canadian Navy, operating out of ports such as St. John’s and Halifax, could, as before, fit under what was essentially a British imperial command structure. This was certainly not true of the U.S. Navy when it entered the conflict in December 1941. Admiral King was known for his keen sensitivities toward the British, and while it might be thought that the United States had so much to do in the gigantic Pacific War that it might find it agreeable simply to leave some of its warships to operate under an Anglo-Canadian command structure in the Atlantic, this did not happen easily. In any case, for much of 1942 the greatest U-boat challenge had occurred in America’s own waters, off its eastern seaboard, and further south, in the Caribbean routes; the U.S. Navy clearly had to be centrally involved here. So, for all those Naval War College lectures about the benefits of having an integrated “command of the sea,” the Allied navies had decided to settle for three zones with identified handover points.
This position improved significantly in early March 1943 following the Atlantic Convoy Conference in Washington. What could have been a serious crisis among the Allies—King for a while wanted to withdraw all U.S. warships from the North Atlantic in order to protect the military supply routes to American forces fighting in Tunisia—ended in a sensible compromise. The U.S. Navy would have primary responsibility for the convoys to Gibraltar and North Africa and would also protect all Caribbean convoys, while the British and Canadian navies assumed responsibility for the major routes to the United Kingdom. More important still was that King agreed to lend some naval forces (including a new escort carrier) to the North Atlantic theater, and did not oppose increasing the numbers of squadrons of very long-range B-24 aircraft to RAF Coastal Command and to the fast-growing but overstretched Royal Canadian Air Force. The last additions, as we shall see, came just in time.
The necessarily important factor of intelligence and counterintelligence fitted well into these larger command structures. Older forms of gaining information about the enemy’s forces and possible intentions still operated in this war, and the British in particular used aerial reconnaissance, reports from their agents and anti-German resistance movements, and technical analysis of captured weapons systems to add to their miscellany of acquired knowledge. Both sides also developed some very sophisticated bureaus of operational research, whose analysts studied runs of data to figure out how best to utilize one’s own resources and dilute the foe’s. But it was in the 1939–45 struggles that signals intelligence, or “sigint,” took a major lead over human intelligence, or “humint,” in the great game of knowing one’s enemy. Nowhere did this seem more important than in the Battle of the Atlantic. For the U-boats to know where the convoys were, or for the Allied navies to glean the disposition of the U-boats, made an utterly critical difference. Small wonder that the code breakers at Bletchley Park included a substantial naval intelligence section reporting directly to the Admiralty, or that Doenitz relied so heavily upon his invaluable B-Dienst.
Nonetheless, the warding off of a submarine attack and the
destruction of the attackers had to be done through technology, that is, by defensive and offensive weapons platforms. It was true, obviously, in all theaters of war and at all times, but it is astonishing how much the exigencies of total war, and the terrible importance of winning the Battle of the Atlantic to both sides, led in the single year of 1943 to a staggering increase in the number of new ways of detecting an enemy and of deploying new weapons to kill him. This conflict was, more than any other battle for the seas, a scientists’ war.
But the acquisition of newer technologies to detect and beat off, or to pursue and destroy, called in turn for their most efficient application—for significant improvements in tactics and training, both by individual submarines, surface escorts, and aircraft and (especially) by groups of them working together. Here the U-boats had an early advantage. They had all-volunteer crews, with some of their commanders remarkably young yet very capable, and saw themselves as an elite branch. They had a single operational task—to sink, and keep sinking, Allied merchantmen, and then avoid being sunk themselves. For a long time they enjoyed the tactical benefit of Doenitz’s switching their mode of attack to nighttime surface encounters. They also possessed a very robust wireless communications system, so if one U-boat spotted a convoy, the other members of the wolf pack would very swiftly know about it and adjust their locations accordingly. Finally, the targets in question were usually very slow-moving and thus offered repeated chances for attack, so even if the submarines waited until the convoys were in the mid-Atlantic air gap, they still had lots of time.
Against such a strong hand the Allies had initially very few trumps. What limited number of aircraft carriers the Royal Navy possessed (often as few as only three or four) had to be deployed for aerial cover of the battle fleets and the Mediterranean convoys. Those same high-profile tasks also consumed the energies of the flotillas of the speedy fleet destroyers. The convoys were thus protected by a tiny group of smaller, slower, and often nearly obsolete craft, lacking aerial protection in the middle stretches, most of them lacking detection equipment, and armed with what were essentially weapons of the First World War. A U-boat could actually outrun most of the early Allied escorts, at least on the surface, if its commander was willing to accept the risk of being spotted—though of course neither hunter nor hunted could go at full speed in the massive Atlantic storms. Much-improved equipment was promised, and some was in the pipeline, but could it be gotten to Liverpool and Halifax and the air squadrons in time?