Wednesday, February 16, 2011

Statistical confusion – whose troops actually did the fighting in World War Two

I was recently researching how many divisions were in action for which nations, at what time and for how long, during the Second World War: and came up with some astonishing misconceptions. (Coincidentally backed up by a recent readers question about who ‘Frenched’… not a term I am familiar with, but I can hazard a guess at its meaning.)

China for instance had theoretically more than 300 divisions, though in fact most were lucky to have the combat power of a Western battalion, perhaps Regiment if they were one of the best equipped. Some of their best ‘Armies’ might have matched a poor Japanese division… maybe. When Stillwell was assigned to rebuild a more useful force on American lines he felt he might assemble about 30 lightweight divisions out of the resources actually available, with no pretence that any of the end products would actually match a Japanese division in the field (even if the Chinese would have let them fight).

The Eastern Front is also a bit fanciful in this regard. Although some German units started each campaign season at or near full strength, for most of the war the vast majority of divisions on both German and Russian sides were perhaps the equivalent of a Western Brigade or Regiment. Many were far weaker (particularly those of Germany’s ‘allies’). As a rule a Soviet Corps might match a weak German division, but you would probably need a small Soviet Army to match a fully mechanised Western division in combat power.

So talk of the Germans having 200+ divisions on the Eastern Front compared to only 80 facing the West tends to hide the fact that a large majority of the Eastern Front units were undermanned infantry, and a far more significant percentage of the units facing West were mechanised, and often at or near full strength. In sheer combat power, the removal of ten percent of divisions (say 20 divisions) from the Eastern Front to face the Western Allies (happened 3 times – Tunisia/Mediterranean 1942, Sicily/Italy 1943, and France 1944) looks a lot more significant if it involves moving 50% of the available Panzers and 70 or 80% of the high quality, full strength, specially equipped, Paratroop or Mountain or Waffen SS divisions. (Though far more Germans – and their Axis Hungarian, Rumanian, Finnish, etc allies – died on the Eastern front than in the west. See my post here for a discussion of the numbers fallacy on the Eastern Front.)

But the really interesting thing was working out the numbers of Western Allied divisions deployed at any point in the war. I, like most others I suppose, knew that American units were not relevant until late1942, but I assumed they formed a large percentage of units in action fairly quickly after that. Certainly I had subconsciously fallen for the idea that by the time of the D-Day invasion the Americans were providing the bulk of the combat troops for the Western Allies. But apparently that is just another example of letting your pre-conceptions run away with you.

Throughout 1942 British Comonwealth troops were fighting, or seriously expecting to be attacked, in French North Africa, Libya, Egypt, Cyprus, Syria (torn between expecting airborne assault, and preparing to reinforce Turkey if that country was attacked), Iraq and Iran (German invasion from the north was attracting more British troop deployment until after Stalingrad than those facing Japan and Rommel combined), Madagascar (fighting the Vichy French to prevent them from inviting the Japanese in as they had done in Indochina), Ceylon (at the time of the Japanese naval raid that looked like it might prefigure and invasion), India, Burma, outposts of the East Indies, New Guinea, Australia, New Zealand, Fiji, and other Pacific Islands. A total of 30+ divisions in combat, and another 30+ expecting imminent attack. (This does not include yet another 30 odd British and Canadian divisions in the UK.) Apart from the Philippino forces surrendered early in the year, the Americans had a couple of divisions in action at Gaudalcanal after August, one in New Guinea by November, and late in November a few arrived in French North Africa.

In 1943 the Americans managed to get their numbers up to half a dozen divisions at the front in Europe and the same in the Pacific, but still not matching the British or Indian armies respectively, and barely matching the combined efforts of minor allies like the Free Poles, French, Greeks and Italians etc.

The breakthrough in American numbers was not until after the middle of 1944, when American units started arriving direct to France (which admittedly, was what Marshall had been trying to do all along).

But although American troops may have outnumbered British and Commonwealth troops in France by late 1944, the total of Allied troops, including the Free French, Poles, Czech, Dutch, Belgians, ensured that it was never quite as clear cut a domination as it appears. Devers ‘American’ 6th Army Group that come up from the South Coast was half French after all. In fact in 1945 it became a race to see if the Americans could import new divisions faster than the French could commission theirs (France had 1.3 million men in the field by VE Day). But the Americans fielding 60 divisions in France compared to only 20 British Commonwealth/ minor allies is the figure waved around as significant. (Ignoring that 15 of the American divisions did not get there until 1945, and by the end the liberated French had mobilised a couple of dozen divisions too, making the non-American total more like 40).

So the Americans did predominate in France, but the war was spread a bit further than France. If you take Europe as a whole, then the situation gets more interesting. The Americans in combat in Europe possibly didn’t start to outnumber the total other Western Allies until about the time of the collapse of Germany’s frontiers, and only weeks before the final surrender.

In Italy American troops never played more than a subsidiary part to the operation, and throughout the war even the ‘American’ 5th Army usually had as many (if not more) British, Canadian, New Zealander, Polish, Italian or French troops in it than Americans. Again, it was not until almost 1945 that even the 5th Army was majority American. They rarely made up more than a third of Allied ground forces in Italy.

If we include the Mediterranean/North African/Middle Eastern forces fighting the ‘anti-German’ half of the World War in a combined ‘European Theatre’ (which was one American generals fanciful suggestion when they wanted Marshall in charge of all ‘European’ operations), then American troops do not dominate ever. There are just too many British and French and Polish and Canadian and New Zealand and South African and Indian and Italian and Greek and Brazilian and other troops garrisoning recently liberated places from Morrocco to Iran and Ethiopia to Belgium; and still fighting to secure Greece, Austria, Denmark and Norway. (Note: The Soviets were starting to pile on pressure in Iran and throughout the Middle East already, and Greece was in serious danger of falling behind the Iron Curtain until British troops did some hard fighting.)

The war against Japan is even more deceptive, particularly if you fall for the fantasy that it was a ‘Pacific’ war. Leaving aside the supposed millions of Chinese, the British Empire and Commonwealth already had more than a million men at the front in India, Burma, Malaysia, New Guinea, Indonesia, and in the Pacific Islands, before the Americans had introduced more than a few divisions. Again, it is almost 1945, less than 10 months before the Japanese surrender, before the Phillipines campaign actually saw an entire American army (the 6th) deployed at a single time, instead of just a division fighting on this island for a month, and two or three on that for a few months. Until well into 1943 the Australian Army alone deployed more ground fighting troops against the Japanese than the Americans. The Americans never put more troops into combat against the Japanese at any point than just the Indian Army (which had a total of 32 divisions at its height, several in Europe or the Middle East, but many of which eventually faced Japan).

On a worldwide scale, the point at which the Americans fielded more troops than just the other Western allies (leaving aside the Russians and Chinese, the Hungarians, Rumanians, Yugoslavs, and all the others who fought the Axis), was… well never. The British Commonwealth alone fielded over 100 divisions in 1942 (though admittedly many were weaker garrison forces than proper mechanised field divisions), compared to the American total of 88 by the end of the war. The French had fielded 100 in 1940, and were to field 20+ again just in France by the end of the war. In fact the largely forgotten minor allies, the Free Poles, the Free Italian combat Groups, the Brigades of Free Greeks, Belgians, Dutch, etc, and the South African divisions, the New Zealand divisions, and the Brazilian division, had between them outnumbered the total American commitment to combat in Europe before the last four months of 1944. Add in the British, Canadians and Free French, and the American commitment before mid 1944 looks rather less impressive than is justified by the hype.

I will even go as far as quoting the figures, taken mostly from John Ellis’ World War two - A Statistical Survey, with a little reference to the microfilm archives of the CCOS deployment figures. (Though I foresee problems with comparing apples and oranges, so please do not consider these numbers as more than a very rough calculation. Particularly as some units have to be estimates. The British Commonwealth uniquely deployed ‘independent armoured brigades’ with roughly the same tank strength as most American armoured division, or some German Panzer Corps, or Russian Tank Armies, which I have accepted in John Ellis’ category and loosely called ½ a division. The same goes for the Italian ‘combat groups’ which I have also ranked as half a division. Many Pacific islands were invaded by a couple of American Regiments, which again could be loosely considered ½ of a division. When I say ‘rough’ estimates, I really mean it.)

The United States divisions were ‘deployed overseas’ for a total of about 1,150 months. Of that: Infantry in Europe about 500, infantry in the Pacific 312, armour 158, marines 128, airborne 37 and cavalry 19… roughly. But ‘deployed overseas’ is a bit different from everyone elses ‘in combat’ definition. For instance US 82nd Airborne is listed in Europe for 19 months from July 1943 to May 1945, but it was out of combat more often than in during that time. By comparison the British 6th Airborne, which was also ‘in Europe’ for all those months, gets listed as actually being in combat for three operations – June - September 1944 for D-Day, December - January 1944 for The Bulge, and March 1945 for The Rhine - and only gets credited with 6 months in combat.

This sample is much worse in the Pacific, where more than 20 American divisions are listed as ‘in Pacific’ for several years, regardless that usually only one or two were actually fighting anywhere at any given time. 1st US Marine Division for instance, probably the hardest fighting US dicvision in the Pacific, is listed ‘in theatre’ for 37 months, August 1942 – August 1945: but apparently fought on Guadalcanal for about five months, then on Cape Gloucestor in New Britain between 26 December 1943 and 16 January 1944 (call it two months?); then on Pelelui for a month, and on Okinawa for three months. Total 11 months, or a bit less than 30% of time 'in theatre' actually in combat.

So compared to a grand total of 1,150 months ‘overseas’ for all American divisions of all types, make what you will of these numbers, all months actually ‘in combat’:

Infantry divisions - British 284 months in combat, Indian 282, Australian 183, Canadian 44, African empire troops 68, South Africa 33, New Zealand 35 (Commonwealth total 935 months in combat). Also Free French 75, Free Poles 34, Free Italians 28, Brazilians 10 and Free Czechs 6, + Greeks, Jews (Palestinian Jews), etc. (Total of minors 153+). Total of just the infantry divisions of the non American Western Allies comes to almost 1,100 months in actual combat. (Although the Americans come up with almost 500 months ‘in Europe’, and 312 ‘in Pacific’, it would be extraordinarly generous to suggest that the total number ‘in combat’ came to more than 60% of that. In real terms it is unlikely that the American total in combat came to half of everyone elses 1,100 months.)

How about armour? British armoured divisions/brigades 245 months ‘in combat’, Indian 18, Australian 25, Canadian 31, New Zealander 9, Free French 27, Free Poles 18, Free Czechs 6. (Total 379 months in combat.) American armoured divisions 158 months ‘in Europe’. Again, even being hugely generous, the American total ‘in combat’ is unlikely to be much more than a third of everyone elses.

(By the way I think the Australian and New Zealand numbers in the Pacific theatre are as woolly and questionable as the American ones, but their African/European numbers are definitely correct, and I think the point is adequately made.)
Total non-American Western Allies army troops in combat about 1,500 months. Somewhere between two and three times total American Army and Marines combined.

Now I am not suggesting that the Americans didn’t contribute. They contributed an awful lot. By the end of the war they contributed more fighting divisions than any one of these named nations (finally equalling the combined total of the reduced numbers of full strength units deployed by the British Comonwealth). But over the total course of the war the United Kingdoms of the British Isles alone had more divisions actually at the front for more combat months than the Americans, as indeed did the French Army before their collapse in 1940… In fact India and Australia combined probably put in more divisional combat months than the US, and throwing in either the South Africans, or the Canadians, or even the New Zealanders, let alone all of them, would make it a certainty. (The Americans should be grateful that the Poles collapsed within a few weeks in 1939, because otherwise they too would have contributed more to the total divisional combat effort in the war than the Americans in Europe too. 47 divisions/brigade groups for – lets give American style generosity and call it 2 months each in the 1939 campaign – plus 127 months later by British or Russian aligned forces thereafter, for a total of 221 months.)

[I would be really interested to see if anyone can provide good evidence against any of these numbers. There must be some other good sources out there?]

Nor am I suggesting that the war could have been won without the Americans… though the total troop numbers do make it seem a far closer concept than most pretend. (And I should note that the American ‘in theatre’ concept would make the comparisons ridiculous if it was equally applied to everyone else. More British and Indian divisions were deployed in Iraq and Iran and ready to go to Turkey in 1942 – just in case of the very real threat that the Germans would break through the Soviets at Stalingrad – than the Americans had ‘overseas’ that year, or indeed the next. If you added all the troops waiting for an invasion of Britian in 1940-41, or Ireland, or Iceland; or Cyprus in 1942, or Syria, or Persia, or India, or Madagascar, or Ceylon, or Australia or New Zealand: the British Commonwealth numbers ‘in theatre’ jump to over three times the total American time ‘overseas’.)

I am suggesting that total American contribution to ground combat is vastly exaggerated by most of the literature. Through the war as a whole it amounted to about a quarter of the Western Allied total all up. Until mid to late 1944, the American contribution was minimal, and could have been replaced with other troops. In Europe their contribution really became important starting in June 1944, and in Asia starting November 1944. (But by 1944 there were more French and Italian and Indian and Polish volunteers than could be trained and equipped, so an idle side thought is that perhaps a lot of this American manpower might have been more valuably deployed as an arsenal of democracy workforce from 1942 - 1945, rather than spending years in training as infantry divisions that only got into action in 1945?) It was not until the end of 1944 when the large majority of American divisions started to make their presence felt worldwide (well, Northern Europe and the Pacific at least, if still not the Mediterannean, Middle Eastern or mainland Asian theatre’s)… at about the time when the European battles were mostly won, when Germany was already falling apart, and when Japan was trying to get the Soviet Union to be a go between in surrender discussions.

As usual, the problem is beware of statistics. Impressive sounding numbers of divisions do not necessarily relate to an actual combat value, particularly if they are not often in action. In terms of contributing to winning the war Chinese ‘divisions’ were a joke, Russian ‘divisions’ were an exaggeration, and the vast majority of American divisions were too late to see fighting in the critical years – early 1942 to late 1944 – when the tide was turned.