. 5
( 14)


1937“8. Inevitably there were teething problems, and production was
often held up by the need to make modi¬cations. Nevertheless, the
speed with which biplanes disappeared from squadron service was
remarkable. In June 1937 only four out of ¬fty-seven bomber squadrons
were completely equipped with monoplanes, although another eleven
had begun to convert to monoplanes; by September 1939 all of the
biplanes had been replaced.47 On the other hand, this rapid moder-
nisation was brought about by producing large numbers of machines
that could not, in terms of range or striking power, properly be called
strategic bombers. The Bristol Blenheim and the Fairey Battle, both
introduced into service in 1937, had too limited a range to reach
the Ruhr from England. The Battle was really an unsuccessful experi-
ment in producing what was then termed a medium bomber with one
engine rather than two. However, the need for political reasons to
attempt to maintain numerical parity led to it being ordered in large
numbers even although it was, in the words of the of¬cial history of
design and development, ˜an aircraft that nobody much wanted™.48
Indeed, it continued in production until December 1940, although the
Air Staff had known in 1936 that it was unsuitable for its purpose. As
late as September 1939 only twenty-eight out of ¬fty-¬ve bomber
squadrons had Whitleys, Hampdens or Wellingtons. All other machines,
including all in service prior to 1937 and the majority down to 1939,
were pseudo-strategic bombers that could not ful¬l the Air Staff™s
The use of RAF aircraft in the imperial policing role might have been
expected to encourage the development of tactical bombers and doc-
trine, but this was not the case. Imperial policing was generally carried
out by slow, general purpose machines that had to be rugged and reli-
able for operations over deserts and mountains, although the Hart,
designed as a fast day bomber for Europe, proved to be suitable for
India.49 The Hart in its day would have been an excellent tactical
bomber, as would its intended successor, the Merlin-powered mono-
plane Hawker Henley light bomber of 1937. However, the Henley,
although faster and more agile than the similarly powered but lumbering
Battle, was relegated to the modest role of target-tug for training anti-
aircraft gunners. Air Staff doctrine was based on strategic bombing,
and therefore the machine with the larger bomb load was preferred.
Chris Ashworth, Bomber Command 1936“1968 (Yeovil: Patrick Stephens, 1995), pp. 21,
Postan, Hay and Scott, Design and Development, p. 11.
Omissi, Air Power, pp. 138“49.
Retrenchment and rearmament 115

Co-operation with the army was left to short-range reconnaissance
machines and RAF bomber crews were not trained to intervene directly
on the battle¬eld.50
The long distances involved in imperial defence encouraged the
development of ¬‚ying boats for reconnaissance, and arguably the RAF
was a leader in this ¬eld. However, co-operation with the navy was not
high on the Air Staff™s list of priorities, and it did not develop an anti-
submarine doctrine or effective aircraft for what became Coastal
Command. Some degree of co-operation with the navy was necessary
regarding the equipment of the Fleet Air Arm, at least until its transfer
from Air Ministry to Admiralty control in 1937, but this co-operation
did not extend to agreement on how best air power could be applied to
naval warfare. The bomber versus the battleship controversy, the classic
inter-service dispute, is dealt with below in connection with naval
armaments. It may be noted here that the RAF™s case would have been
more convincing if it had provided some of its long-range bomber crews
with the specialised training required to attack warships, but this role
was left to small numbers of specialised torpedo bombers, either carrier-
or shore-based.
British warplanes did not always ful¬l their potential or the Air Staff™s
hopes, but problems tended to arise from a mismatch of strategic or
tactical doctrine with technical possibilities rather than with short-
comings in aircraft design. The traditional view that the inter-war British
aircraft industry was backward has been contested by Edgerton, who has
pointed to its military focus and the support it received from govern-
ment. The Air Ministry was easily the largest research and development
institution in the country and there were great advances in aerodynamic
theory as a result of work in the universities and at the government™s
Royal Aircraft Establishment.51 The fact that biplanes, on the eve of
rearmament, had a super¬cial resemblance to First World War aircraft
does not justify Barnett™s description of the products of the inter-war
British aircraft industry as ˜backward in design and construction™,
employing wood and fabric when America and Germany had already
moved into the era of all-metal aircraft.52 In fact the Bristol aircraft
company produced an all-metal, monoplane ¬ghter, the Type 133, with
retractable undercarriage in 1934, a year ahead of the prototype of the
Messerschmitt Bf 109.53 The RAF chose the Gladiator biplane in 1934

Sir John Slessor, Central Blue (London: Cassell, 1956), pp. 166, 183, 204“8.
Edgerton, England and the Aeroplane, pp. 18“37.
Barnett, Collapse of British Power, pp. 213, 418.
Peter Lewis, The British Fighter since 1912 (London: Putnam, 1979), p. 232; William
Green, Warplanes of the Third Reich (London: MacDonald and Jane™s, 1970), p. 526.
116 Arms, economics and British strategy

because it could be in service sooner than the Bristol machine. Reten-
tion of the biplane format, with its greater manoeuvrability, if lower
speed, was not exceptional: the Italian and Soviet air forces both had
biplane ¬ghters until 1942, and even the Luftwaffe had one of its thirteen
day ¬ghter wings still equipped with biplanes in September 1939.54
British aircraft designers made use of metal from an early date, the
change from wood to metal airframe having begun with the Armstrong
Whitworth Siskin ¬ghter, which entered service in 1924, and fabric-
covered metal frames were standard for German as well as British
¬ghters in the early 1930s.55 In any case, the move from wood to metal
did not necessarily represent progress: the bias of American aircraft
designers in favour of metal precluded promising projects involving
wood, such as the de Havilland Mosquito of the Second World War,
which drew upon that ¬rm™s experience of using wood in the 1930s.56
British aero-engine ¬rms, particularly Rolls-Royce and Bristol, were so
far from being backward in 1935 that the prototypes of the Mes-
serschmitt Bf 109 ¬ghter and the Junkers Ju 87 Stuka dive bomber were
¬tted with Rolls-Royce engines in that year because German ¬rms had
lagged behind their British counterparts.57
One notable aspect of the revolution in aircraft design in the mid-
1930s was a sharp increase in the prices of all types. Table 3.1 is based
on estimates made by the Air Ministry in 1937 and although the ¬gures
are expressed in current prices they are reasonably comparable, since
retail prices were generally lower in 1937 than in 1924.58 As light
bombers like the Hart and Hind were replaced by larger machines like
the Battle and Blenheim in 1937, the unit cost rose by about three times.
The larger medium bombers of 1939, the Hampden and Wellington,
were expected to be more expensive than earlier heavy bombers. Nor
was a slackening in the increase in costs in sight. The four-engined
bombers like the Halifax that the Air Staff wished to order in 1939 were
estimated to cost 40 per cent more than the current twin-engined
heavy bomber, the Whitley. In contrast, unit costs of both aircraft and

John Killen, The Luftwaffe: A History (London: Frederick Muller, 1967), p. 95.
A. J. Robertson, ˜The British aircraft industry and the state in the interwar period: a
comment™, Economic History Review, 28 (1975), 648“57.
Eric Schatzberg, Wings of Wood, Wings of Metal: Culture and Technical Choice in
American Airplane Materials, 1914“1945 (Princeton University Press, 1999).
Green, Warplanes of the Third Reich, pp. 42“4, 168“9, 428, 526.
The nearest thing available to a retail price index in the inter-war period, the Ministry of
Labour™s working-class cost of living index, is not a good price de¬‚ator as it was based
on Edwardian consumption patterns. More signi¬cant regarding relative prices of
aircraft is the fact that average money wages “ the major input “ were stable from 1924
and were no higher in 1936 than in 1929.
Retrenchment and rearmament 117
Table 3.1. Costs of aircraft, 1924“39 (£ at current prices)

1924 Siskin 4,050
1930 Bulldog 4,200
1934 Fury 3,900
1937 Gladiator 5,300
1939 Spit¬re 8,000
1939 De¬ant 10,500
Light bombers
1924 DH 9A 3,300
1930 Hart 4,050
1937 Hind 4,150

Medium bombers
1937 Battle 11,250
1937 Blenheim 13,500
1939 Hampden 20,000
1939 Wellington 18,500

Heavy bombers
1924 Virginia 13,300
1934 Heyford 13,750
1937 Whitley 30,000
1939 Halifax 42,000

Sources: ˜Defence expenditure in future years: summary of forecasts submitted
by the defence departments™, 22 Oct. 1937, CAB 24/272, TNA.

warships had been relatively stable between the mid-1920s and the
mid-1930s: for example, a biplane ¬ghter cost 3.7 per cent less in 1934
than in 1924, and a class of cruisers of about 7,000 tons completed
between 1933 and 1935 cost 8.8 per cent less on average than a class of
7,500-ton cruisers completed in 1926.59

Naval weapons
In 1919 the Naval Staff argued that aircraft were in their infancy and
that the battleship could not be dispensed with for the foreseeable
future: ˜the country whose fast capital ships and their complementary
units are not contained or held by similar enemy ships can, with these
vessels, sweep the enemy ships and sea-borne trade off the sea™.60

Data for cruisers from Jane™s Fighting Ships 1936 (London: Sampson Low, Marston &
Co, 1936), pp. 51“2.
˜Some notes by Naval Staff for First Lord™s speech on the naval estimates: the retention
of capital ships™, 24 Nov. 1919, Admiralty records, series 116, ¬le 1677 (ADM 116/
1677), TNA.
118 Arms, economics and British strategy

Complementary units included aircraft carriers. Naval historians have
tended to assume that the Admiralty was irrationally attached to bat-
tleships and allowed Britain to fall behind in naval aviation. Edgerton,
however, has pointed out that in 1935 Britain had six carriers compared
with four each for the United States and Japan, and would have had
more than either of these powers at the beginning of the Paci¬c War in
1941 had it not been for wartime losses.61 The lead over European
powers was much greater: in 1939 France had only one ¬‚ush-decked
carrier, which was too slow for a ¬‚eet action, and the German and
Italian navies had none in service. The Admiralty™s commitment to
naval aviation was indicated by the fact that as many aircraft carriers
(¬ve) were launched between 1937 and 1940 as battleships. Opposition
to building carriers came not from conservative admirals but from the
Air Ministry, which claimed that land-based aircraft would be better
value for money. This interdepartmental dispute gave the Treasury the
opportunity to delay the laying down of the ¬rst of the new carriers, the
Ark Royal, until 1934, when the First Lord, Sir Bolton Eyers-Monsell,
was provoked by the Air Ministry™s claim to remark that ˜six-inch guns
ashore would cost only a fraction of the cost of a ship mounting guns of
the same calibre “ but we cannot on that account do without the ship™.
The Treasury conceded that the Admiralty had got the better of the
The inter-war bomber versus the battleship debate in Whitehall was
largely, although not entirely, an interdepartmental dispute over
resources at a time when the Treasury was trying to reduce the annual
estimates. In November 1920 the Admiralty estimated that the con-
struction of four new capital ships would cost £37,500,000.63 In the
event, these 48,000-ton monsters were cancelled under the Washington
naval treaty and replaced by two 35,000-ton battleships. Even these
˜Nelson™-class vessels cost £7.5 million each, at a time when the naval
estimates were being reduced from £80.8 million in 1922/3 to £52.6
million in 1924/5. The Air Ministry™s estimates in 1922/3 amounted to
only £9.4 million. The Admiralty™s request to proceed with new capital
ships in 1920 prompted an enquiry by a CID sub-committee on Naval
Shipbuilding, chaired by Bonar Law. Most of the witnesses agreed that
the development of aircraft had not yet reached a stage where the bat-
tleship was obsolete. Bonar Law attached particular importance to
Trenchard™s evidence that aircraft were still ineffective against a capital
Edgerton, Warfare State, pp. 30“2.
Eyers-Monsell to Sir Philip Sassoon (Air Ministry), and notes by Bridges, 4 Oct. and 11
Oct. 1934, T 161/624/S.36130/34, TNA.
˜Naval policy and construction™, CP 2176, 22 Nov. 1920, CAB 24/124, TNA.
Retrenchment and rearmament 119

ship, although the CAS clearly believed that the air threat would grow.
Naval advocates of torpedo bombers and carriers argued that research
into these weapons would be a better investment than construction of
battleships. Such views were stoutly opposed by Chat¬eld, then assis-
tant chief of the Naval Staff, who argued that ships with armoured decks
designed to withstand plunging 16-inch shells from other capital ships,
and ˜bulges™ to absorb torpedoes ¬red from submarines, would resist
any bombs carried by aircraft.64 The enquiry did not accept the radical
case against the battleship, but trials were carried on from 1921 into the
effectiveness of armour against bombs, and ˜bulges™ against near misses
by bombs or aerial torpedoes, and into the accuracy of bombing (using a
radio-controlled, but slow-moving old battleship) and of anti-aircraft
guns (using slow, unmanned target aircraft), without producing an
interdepartmental consensus. The debate only acquired urgency in
1936, when the moratorium on capital ship construction, agreed at
Washington and extended in 1930 at the ¬rst London naval conference,
was about to end.
In March 1936 a CID sub-committee was appointed to consider the
vulnerability of capital ships to air attack, in the light of experiments
since 1921, with Inskip in the chair, and Chat¬eld, now ¬rst sea lord,
and Ellington, the CAS, as his expert advisers. Chat¬eld™s view, as in
1921, was that no warship was unsinkable but there was no reason why
capital ships could not be designed to withstand air attack. Ellington
thought there was a tendency in the sub-committee™s discussions for the
capital ship of the future to be pitted against the aircraft of today,
whereas the air threat would increase very rapidly. On the question of
relative price, the Admiralty and the Air Ministry agreed that forty-three
twin-engined medium bombers could be built and maintained for the
cost of one capital ship over a given period (aircraft would have had to
be replaced more frequently). Inskip did not accept, however, that such
aircraft could be substituted for battleships. Given the need to protect
Britain™s world-wide trade and empire, there was a need for ships equal
in ¬ghting power to the enemy that could operate in the oceans far
beyond the range of land-based aircraft. He felt that Britain had more to
lose by making a wrong decision in this matter than other naval powers,
yet none of them proposed to dispense with battleships. He accepted
that it was extremely dif¬cult to conduct tests in peace that would be
suf¬ciently realistic to be conclusive, but urged that the Admiralty and

Geoffrey Till, ˜Airpower and the battleship in the 1920s™, in Bryan Ranft (ed.),
Technical Change and British Naval Policy 1860“1939 (London: Hodder and Stoughton,
1977), pp. 108“22, at pp. 111“13.
120 Arms, economics and British strategy

Air Ministry should collaborate in further trials aimed at obtaining
information that would enable capital ships to be designed so as to
secure the maximum immunity from air attack.65
Given uncertainty whether all other naval powers would observe the
35,000-ton limit ¬rst set at Washington in 1922, and renewed by
Britain, France and the United States at the second London naval
conference in 1936, there was doubt whether the new capital ships due
to be laid down in January 1937 would be equal in ¬ghting power to a
potential enemy. The signatories of the second London naval treaty
agreed on 14-inch calibre as the maximum size of gun, but the Japanese
refused to state their intentions and the Americans then opted for a
16-inch gun. However, the British had already designed mountings for
14-inch guns and, as the construction of gun mountings determined
how soon a capital ship could be completed, Chat¬eld decided to go
ahead with 14-inch guns for the ˜King George V™ class. Speed of con-
struction seemed to be of the essence, given that all except three of the
Royal Navy™s capital ships were at least twenty years old, and the fact
that France, Germany and Italy were already building new ones. The
completion of the ˜King George V™ class was, however, delayed by
Chat¬eld™s insistence on extra armour, which necessitated the redesign
of one of the turrets to take two rather than four guns, to save weight.66
The consequence was that Britain™s new capital ships had smaller calibre
guns than those constructed by the Italians, Germans or Japanese, but
this was probably inevitable, given that the British observed the 35,000-
ton limit and the Germans and Japanese did not.
Cruisers were considered by the Naval Staff to be the major threat to
British trade, and also the means by which a guerre de course could be
conducted against Japan. The Germans sprang a surprise in the early
1930s when they began to replace their pre-dreadnoughts with so-called
pocket battleships. These armoured ships were nominally within the
10,000-ton limit set by the Treaty of Versailles, but in reality displaced
20 per cent more. They were designed as commerce raiders, with suf-
¬cient speed to evade all capital ships, except battle-cruisers, but with
more powerful guns (11-inch) than any allowed to cruisers under the
Washington treaty (maximum 8-inch). However, the Admiralty gave
greater priority to the Japanese menace. When asked in the DRC in

˜Sub-Committee on the Vulnerability of Capital Ships to Air Attack, report™, CID paper
1258-B, 30 July 1936, and minutes of evidence, CAB 16/147, TNA. An abridged
version of the report was published as Cmd 5301 (PP 1936“7, xii. 73“90).
Chat¬eld to Churchill, 10 Mar. 1942, Chat¬eld Papers, CHT/4/3, National Maritime
Museum, Greenwich; G. A. H. Gordon, British Seapower and Procurement between the
Wars: A Reappraisal of Rearmament (Basingstoke: Macmillan, 1988), p. 173.
Retrenchment and rearmament 121

January 1934 by Hankey and Fisher about the pocket battleships,
Chat¬eld replied that the French (who had laid down the ¬rst of two
battle-cruisers in 1932) could look after them. He added that Britain™s
three battle-cruisers could do so too, but ultimately the Royal Navy
might not possess any ships of that type as the design of their replace-
ments would depend upon what the Japanese replaced their battle-
cruisers with.67 No attempt was made to build larger cruisers to cope
with the pocket battleships. Indeed, the Admiralty proposed at the
second London naval conference to reduce the maximum size of cruisers,
from the 10,000 tons allowed under the Washington treaty to 8,000
tons, to make it possible to build at less cost the number believed to be
required to protect British trade.68
The need for economy also in¬‚uenced Admiralty thinking on anti-
submarine warfare. On being appointed ¬rst sea lord, Chat¬eld made a
start to building a reserve of anti-submarine vessels, equipment and
trained personnel to form a nucleus for the anti-submarine force that
would be required in war. However, in a period of severe ¬nancial
restraint his only change in the Admiralty™s sketch estimates for 1933
was to raise the number of anti-submarine escorts to be laid down from
one to two.69 When Fisher asked in the DRC in January 1934 whether
the navy had enough destroyers for anti-submarine work, Chat¬eld said
that the number could be increased by refraining from scrapping old
vessels, if the Germans started building submarines again. He also
declined to take up Fisher™s suggestion that the Admiralty should put in
a claim for all the Asdic (Sonar) sets required in war, instead of asking
for 200 and waiting for others to be manufactured subsequently.70
Germany had already taken secret steps to re-create a submarine force
and the Admiralty was not entirely surprised to learn that U-boats were
already under construction when Hitler denounced the arms limitation
clauses of the Treaty of Versailles in March 1935. Even so, the
Admiralty continued to give a higher priority to major warships than to
anti-submarine vessels, on the grounds that the latter could be produced
quickly in an emergency. By 1934 the Admiralty had designed an anti-
submarine trawler capable of rapid construction and in October 1938 it
was decided to build a new type of ship that would be cheaper
than conventional destroyers, but faster than existing designs of escort

DRC minutes, 23 Jan. 1934, CAB 16/109, TNA.
Stephen Roskill, Naval Policy between the Wars, 2 vols. (London: Collins, 1968“76),
vol. II: The Period of Reluctant Rearmament 1930“1939, p. 317.
Chat¬eld™s comments on ˜Naval policy and the sketch estimates for 1933™, 30 Aug.
1932, CHT/3/1, National Maritime Museum, Greenwich.
DRC minutes, 18 Jan. and 23 Jan. 1934, CAB 16/109, TNA.
122 Arms, economics and British strategy

vessels.71 The Royal Navy was ahead of the US Navy in ¬tting Asdic to all
destroyers from 1932 as well as to older ones earmarked for escort duties.
The development of the Royal Navy™s own submarines was restricted
by the need for economy, although insofar as economy led to standar-
disation of design it was not without bene¬t. There had been ¬ve dif-
ferent kinds of submarine in service in 1918: patrol, ¬‚eet, minelayer,
monitor and anti-submarine. The last two types were quickly dis-
continued and, although examples of the other types were still in service
in 1939, only the patrol type was under continuing development.
Increasing concentration on patrol submarines capable of laying mines
as well as reconnaissance and anti-warship patrols was the best use of
scarce resources.72
On balance, it is clear that the Royal Navy did not cease to be a
technical leader in the inter-war period. Geoffrey Till has argued that
the British were less perceptive than the Japanese or the Americans
in their understanding of what naval aviation could achieve, but, as he
says, the difference was only one of degree.73 In other respects the Royal
Navy was clearly ahead. The development of Asdic had been kept secret
even from Allied navies in the First World War and Britain still had a
strong lead in 1939.74 The Royal Navy was also participating in the
development of radar from 1935, and the ¬rst sets were installed in a
cruiser in 1938. Britain was thus well placed to exploit electronic
technology, which came to dominate naval warfare in the Second World
War, both as regards tracking an enemy (as with the shadowing of the
battleship Bismarck in 1941) and ¬ghting at night.75

Army weapons
Britain continued to lead the world in the development of tanks during
the 1920s. The Vickers Medium Mark I, deliveries of which began in
1923, was the ¬rst British tank in service to have a revolving turret. The
Mark III of 1928 was well ahead of its time, having the ¬repower of a
heavy tank but the speed of a medium one. However, because of its high

First Sea Lord, Sir Roger Backhouse, to Inskip, 13 Oct. 1938, ADM 205/3, TNA.
David Henry, ˜British submarine policy, 1918“1939™, in Ranft (ed.), Technical Change
and British Naval Policy, pp. 80“107.
Geoffrey Till, ˜Adapting the aircraft carrier: the British, American and Japanese case
studies™, in Williamson Murray and Alan R. Millett (eds.), Military Innovation in the
Interwar Period (Cambridge University Press, 1996), pp. 191“226.
George Franklin, Britain™s Anti-Submarine Capability 1919“1939 (London: Frank Cass,
Alan Beyerchen, ˜From radio to radar™, in Murray and Millett, Military Innovation,
pp. 265“99, at p. 294.
Retrenchment and rearmament 123

cost, only three examples were built and the type was not developed.
Instead the army concentrated on light tanks, initially of 4.25 tons, little
more than a quarter of the size of the Medium Mark III. Armed only
with machine guns, they were of no use in tank-to-tank combat, but they
could be built cheaply and issued for training. An identical policy was
followed by the German army when it adopted the Panzerkampfwagen 1,
with similar characteristics to the British Light Tank Mark VI, in 1934.
Reliance upon light tanks was not a sign of technical backwardness.
Indeed, British tanks were widely exported or produced under licence
abroad in the inter-war period; for example, Italy based its standard light
tank down to 1941 on a British model of 1929.
Whereas the United States abolished its tank corps at the end of the
First World War, the Royal Tank Corps (RTC), as it became in 1923,
survived and engaged in a good deal of experiment. Indeed, until the
early 1930s the British army was widely regarded as the world leader in
developing doctrine for tank warfare. This reputation was not entirely
deserved. For example, Fuller, who was the best-known British radical
military thinker of the 1920s, successfully converted the RTC to the
doctrine of ¬ring on the move in tank-to-tank combat. The Germans
adopted the opposite tactic of ¬ring when stationary, which made for
much greater accuracy. From 1927 the British army conducted
experiments to discover how tanks could best be combined with
armoured cars, motorised machine guns and artillery, and aircraft, but
the conclusion drawn was that tanks were best employed in independent
brigades, rather than in units combining different arms, in contrast to
what the Germans were to do with their panzer divisions.76 By late 1933
the British army had an armoured brigade as part of its establishment,
but the composition of the brigade had yet to be decided, pending
experiments with new medium tanks, especially as regards their resis-
tance to anti-tank guns. The cavalry division was still mounted on horses
and would remain so, in the words of the DRC report in February 1934,
˜until a vehicle is designed capable of replacing the horse, and no such
vehicle is in sight™.77 The vulnerability of horsed cavalry to modern
¬repower was recognised, but tanks were as yet characterised by short
range and limited mechanical reliability, and armoured cars lacked
cross-country capability. Nevertheless, in October 1934 the CIGS,
Montgomery-Massingberd, decided that one of the army™s two horsed
cavalry brigades should be given tanks, and combined as a mechanised

Harris, Men, Ideas and Tanks, pp. 197“204, 215“29, 246“8.
˜Defence Requirements Sub-Committee, report™, DRC 14, 28 Feb. 1934, CAB 24/247,
124 Arms, economics and British strategy

cavalry brigade with the army™s sole existing tank brigade to form a
˜mobile™ (armoured) division.78 In the absence of any satisfactory
alternative, the mechanised cavalry were equipped with light tanks.
Conversion of other horsed cavalry regiments followed and, contrary to
what many military historians have assumed, the limiting factor was the
availability of armoured ¬ghting vehicles rather than sentimental
attachment to horses.79
Teething troubles and indecision about the best balance between
armour and speed led to hesitation in adopting a successor to the
obsolescent medium tanks dating from the 1920s. Design work was
delayed by the death of Sir John Carden, Britain™s leading tank
designer, in an aeroplane crash in 1935. The limited resources of
Vickers-Armstrong, Britain™s only producer of tanks after it took over
the Carden-Loyd ¬rm in 1928, were further stretched when the War
Of¬ce decided in 1936 on a re-equipment programme based on no
fewer than four categories of tank. These were to be: a light tank for the
cavalry; a larger ˜cruiser™ tank, armed with a 2-pounder shell-¬ring gun,
but fast enough for the light tank role in the Tank Brigade; a similarly
armed, but more heavily armoured medium tank for the ˜hitting™ role in
the Tank Brigade; and a heavily armoured, slow assault tank to operate
with the infantry. The Cruiser Mark I was not ready for production until
August 1937, and was considered to be too vulnerable to anti-tank guns;
the more heavily armoured Mark II was not ready for production until
July 1938. Neither the Mark I nor the Mark II was suitable for mass
production by non-specialist ¬rms and both were regarded as stop-gaps.
The Cruiser Mark III was developed by the Morris Car Company, using
the suspension of the American Christie tank, but suffered from many
teething problems and did not enter production until December 1938.
The medium tank remained a research project. The Infantry Tank Mark
I was not ready for production until April 1937 and, being armed only
with a single machine gun, was not designed for tank-to-tank ¬ghting.
This drawback was overcome with the Mark II, which had a 2-pounder
gun, but the pilot model did not appear until April 1938 and production
did not start until the following year. The British tanks of the late 1930s
were comparable in ¬ghting power to their German equivalents, but not
in mechanical reliability, re¬‚ecting their hurried development.80 Tanks
suitable for use by armoured divisions in a campaign on the European

Minute by CIGS, 15 Oct. 1934, WO 32/2847, TNA.
David French, ˜The mechanization of the British cavalry between the world wars™, War
in History, 10 (2003), 296“320.
Postan, Hay and Scott, Design and Development, pp. 240, 309“13, 316; Harris, Men,
Tanks and Ideas, pp. 240“1; ˜Tank situation statements made to the DPR
Retrenchment and rearmament 125

continent were not going to be available in quantity until 1939 at the
earliest, whatever decisions were to be taken about the army™s role in
1934 or in 1937.
The position about ¬eld artillery was dominated for most of the period
by the huge stocks of guns and ammunition left over from the First World
War. Field artillery units down to the mid-1930s were equipped with the
old 18-pounder and 4.5-inch howitzer, both of which had been out-
ranged by equivalent guns of continental armies and were no longer, in
the War Of¬ce™s view, ˜¬t for war™.81 A speci¬cation had been issued in
May 1934 for a weapon to take the place of both the 18-pounder and the
howitzer, and this led to the excellent 25-pounder used in the Second
World War. Satisfactory trials were held in October 1938 and orders
placed forthwith. Meanwhile existing stocks of 18-pounders were con-
verted to 25-pounders by relining them and modifying the carriage, and
stocks of 60-pounder medium guns were also modernised by being
provided with a loose liner, developments that were described as ˜win-
ners™ by the CIGS, Sir Cyril Deverell, and the Master-General of the
Ordnance, Sir Hugh Elles. The 2-pounder tank gun was adopted as an
anti-tank gun in 1934 and was superior to its German 37-mm equivalent;
however, as early as April 1938 it was recognised that there would
probably be a requirement for a heavier anti-tank gun to cope with
increasing armour, and a new 6-pounder had reached the wooden mock-
up stage by September 1939.82 The artillery and infantry were com-
pletely motorised in the 1930s “ in contrast to continental European
armies, including the German army, where horses continued to be widely
used during the war. The British army was modernised between 1936
and 1939, but the low priority given to it compared with the air force and
navy meant that there continued to be de¬ciencies, especially as regards
guns and tanks, and there was a lack of equipment even for training.83

The economy and ¬nance
What economic and ¬nancial factors prevented Britain from creating
and maintaining an air force equal to any other within striking distance

Sub-Committee, May 1937“May 1938™, WO 332/4441; Army Council “ Informal
Meetings, 13 July and 20 July 1937, WO 163/47, TNA.
˜Organisation, armament and equipment of the army™, DPR 145, 16 Nov. 1936, CAB
64/35, TNA.
Postan, Hay and Scott, Design and Development, pp. 247, 257“9, 261“2, 315“16;
correspondence between CIGS and the Master-General of the Ordnance, 24 Feb.
1937, WO 32/4385; ˜Anti-tank equipment™, WO 32/4684, TNA.
The Ironside Diaries 1937“1940, ed. Roderick Macleod and Denis Kelly (London:
Constable, 1962), pp. 53“4, 56“60.
126 Arms, economics and British strategy

of the United Kingdom? Or maintaining a navy capable of defending
Britain™s interests in the Far East from the 1920s, while also dealing with
the German and Italian navies from 1935 and protecting Britain™s
worldwide trade routes? Or creating an army capable of co-operating
with European allies, while also defending Britain™s overseas territories
and interests? Merely to list Britain™s defence requirements is to show
that not all could be met, and that therefore strategic choices would be
Table 3.2 shows the course of expenditure by the three defence
departments. The ¬gures for the ¬nancial years 1919/20 to 1922/3
include the cost of the aftermath of the war and are not therefore strictly
comparable to the years that followed. Those for 1939/40 include the
cost of operations from September 1939 to March 1940, but these were
limited in scope and in July 1939 the Treasury expected total defence
expenditure to reach £750 million in the current ¬nancial year, even if
there were no war.84 The percentage ¬gures in column 2 show what
share of current national output went to the defence departments, but
when reading them one should remember that GDP at current prices
was affected by a fall in prices in most years from 1920 to 1932, and by
severe falls in output between 1920 and 1921, and between 1929 and
1932. On the other hand, GDP at current prices rose sharply between
1935 and 1937 and again between 1938 and 1939. The percentage
¬gures re¬‚ect changes in the denominator as well as the numerator, and
give an impression that defence expenditure rose more slowly than it did
after 1934. On the other hand, the ¬gures for expenditure in column 1
include the effect of prices rising in response to increased demand for
equipment after 1936. For example, a light tank that cost £3,250 in
October 1936 cost £4,000 in March 1938. Prices on Admiralty con-
tracts increased by about 7 per cent between 1937 and 1938, although
wholesale prices were falling and average money wage rates increased by
only 1.2 per cent between these years.85
Notwithstanding these caveats, table 3.2 shows clear trends. Defence
expenditure stabilised after 1923, even rising in some years and, mea-
sured as a percentage of GDP, was not much lower between 1923/4 and
1927/8 than between 1906/7 and 1913/14 (see table 1.1). From 1928/9
to 1932/3 there was a clear downward trend, with no signi¬cant recovery
until 1935/6, and the percentage ¬gures were stable only because of the

˜Note on ¬nancial situation™, CP 149 (39), 3 July 1939, CAB 23/100, p. 138, TNA.
G. Oram to E. Compton, 23 Mar. 1938, T 161/822/S.35177/38; Captain V. H.
Danckwerts to First Lord, 9 May 1938, ADM 116/3631, TNA; B. R. Mitchell and
Phyllis Deane, Abstract of British Historical Statistics (Cambridge University Press,
1962), pp. 345, 478.
Retrenchment and rearmament 127
Table 3.2. Defence departments™ expenditure as percentage of GDP, 1919/20“1939/40

Financial Total Percentage of GDP
year expenditure (£m.) adjusted to ¬nancial year

1919/20 616.6 12.4
1920/1 232.4 4.6
1921/2 207.4 4.8
1922/3 138.1 3.6
1923/4 122.0 3.3
1924/5 114.7 3.0
1925/6 119.4 3.0
1926/7 116.7 3.0
1927/8 117.4 2.9
1928/9 113.5 2.8
1929/30 113.0 2.7
1930/1 110.5 2.7
1931/2 107.3 2.8
1932/3 103.0 2.8
1933/4 107.9 2.8
1934/5 113.9 2.8
1935/6 136.9 3.2
1936/7 186.1 4.1
1937/8 262.1 5.5
1938/9 382.5 7.7
1939/40 719.0 13.6

Sources: 1919/20 to 1923/4: John Ferris, The Evolution of British Strategic Policy, 1919“26
(Basingstoke: Macmillan, 1989), p. 216; 1924/5 to 1938/9: Statistical Account for the United
Kingdom for Each of the Fifteen Years 1924 to 1938 (Cmd 6232), PP, 1939“40, x, 367;
1939/40: Robert P. Shay, British Rearmament in the Thirties (Princeton University Press,
1977), p. 297.

fall in GDP between 1929 and 1932. While John Ferris™ description of
the period 1921“7 as a peak in British military preparations in
peacetime is something of an over-statement, he is correct in identi-
fying retrenchment between 1928 and 1933 as the main cause of
the de¬ciencies that policymakers had to attempt to remedy from
British power was bound to be affected by the changing fortunes of
the national economy compared with other countries. In the severe
depression of 1921 British GDP fell to 87.1 per cent of the 1913 level
and the 1913 level was not achieved again until 1925. Economic growth

John Robert Ferris, The Evolution of British Strategic Policy, 1919“26 (Basingstoke:
Macmillan, 1989), p. 180; John Robert Ferris, ˜ ˜˜It is our business in the Navy to
command the seas™™ ™, in G. Kennedy and K. Neilson (eds.), Far-Flung Lines.
128 Arms, economics and British strategy

lagged behind other industrial countries from 1913 to 1929, GDP
having risen by only 11.9 per cent over that period compared with
63 per cent for the United States and 78.9 per cent for Japan (two
countries whose economies had bene¬ted from foreign orders during the
war and rapid development during and after it). Even France and
Germany achieved increases of 25.8 per cent and 21.6 per cent
respectively, despite suffering at least as much economic disruption as
Britain during and after the war. There are many reasons for the British
economy™s disappointing performance, including Churchill™s decision,
as chancellor of the exchequer, to return to the gold standard in 1925 at
an over-valued exchange rate; slower growth in international trade than
before 1914; and a reluctance or inability to change more quickly from
an economy based on coal, cotton, iron and steel, shipbuilding and
heavy engineering to one based on oil, man-made ¬bres, motor vehicles,
light engineering and electrical goods. A world depression in the early
1930s saw British GDP fall by 5.1 per cent between 1929 and 1932, but
this loss of output was mild compared with what other countries
experienced: 28.0 per cent in the United States; 11.0 per cent in France
and 15.8 per cent in Germany. Moreover, Britain recovered more
quickly from the depression. Helped by suspension of the gold standard
in 1931 and low interest rates from 1932, she reached her 1929 level of
GDP by 1934, one year ahead of Germany and ¬ve years ahead of
France and the United States.87 However, Germany and Japan
experienced very rapid economic growth in the 1930s and by 1937 the
relative GDPs of the leading capitalist economies compared with 1913
were as shown in table 3.3: Britain had lost ground compared with all of
them except France “ the country with which she was most likely to be
allied in war. Although reliable ¬gures for Soviet GDP are not available,
rapid industrialisation under Stalin meant that Russia was also relatively
more powerful than in 1913.
Falling prices and economic depressions reduced the rate of return
from taxes, yet chancellors of the exchequer faced a major increase in
outgoings as a result of loans raised to pay for the war. The cost of
servicing the national debt (interest, sinking funds and other charges)
increased from £24.5 million, or 12.4 per cent of total revenue, in
1913/14 to an average of £344.5 million, or 37.4 per cent, in the 1920s.
The 1913/14 ¬gure had been about one-third of the combined army and

Maddison, Phases of Capitalist Development, pp. 174“5. For a brief account of why
Britain™s experience differed from other countries, see Barry Eichengreen, ˜The inter-
war economy in a European mirror™, in Roderick Floud and Donald McCloskey (eds.),
The Economic History of Britain since 1700, 2nd edn, 3 vols. (Cambridge University
Press, 1994), vol. II, pp. 291“317.
Retrenchment and rearmament 129

Table 3.3. Indices of GDP for United Kingdom, France, Germany, Italy, Japan and the USA,
1913“37 (1913¼100)

1921 1929 1932 1937

UK 87.1 111.9 106.2 130.9
France 80.5 125.8 112.0 120.8
Germany 87.5 121.1 102.0 153.4
Italy 98.0 130.4 125.5 146.1
Japan 146.9 178.9 173.2 261.2
USA 112.1 163.0 117.4 160.7

Sources: Angus Maddison, Phases of Capitalist Development (Oxford University Press,
1982), pp. 174“5.

navy estimates; the 1920s average was about three times the combined
cost of the peace-time air force, army and navy.88 Fortunately interna-
tional interest rates fell in the depression of the early 1930s and the cost
of servicing the debt was substantially reduced by converting 5 per cent
War Loan to 3.5 per cent in 1932, saving £42 million a year. There was
also a saving of £24.6 million from paying lower rates on Treasury bills.
The decision in 1932 to make only a nominal repayment of the war debt
to the United States, instead of the regular instalments agreed in 1923,
saved a further £24.2 million.89 Altogether these savings on the servi-
cing of the national debt released a total of £90.8 million for other
purposes, a not insigni¬cant sum at a time when defence expenditure
amounted to £103 million.
Unemployment relief was another major burden on the chancellor™s
budget. Unemployment rose in the depression of 1921 to 16.9 per cent
of workers in occupations covered by the National Insurance Act of 1920;
fell to about 10 per cent in the late 1920s; rose to about 22 per cent
in 1931 and 1932, and then fell again to just under 11 per cent in 1937.
Even so, the cost to the Exchequer was higher in the 1930s than the
1920s, as many workers had exhausted their right to unemployment
insurance bene¬ts and relied on Treasury-funded unemployment
assistance. When Sir Richard Hopkins, the second secretary of the
Treasury, was asked in November 1937 about future budget prospects,
and therefore what could be afforded for defence within a balanced
budget, he drew attention to the cost of unemployment assistance as a

Calculated from ¬gures in Bernard Mallet and C. O. George, British Budgets, 3rd Series,
1921/22“1932/33 (London: Macmillan, 1933), pp. 554“7.
Jeremy Wormell, The Management of the National Debt of the United Kingdom, 1900“1932
(London: Routledge, 2000), pp. 622, 685.
130 Arms, economics and British strategy

major consideration, noting that in contrast to the ¬gure of £20 million
in 1929, it did not now fall below £55 million, owing to the numbers of
long-term unemployed. It was capable of rising to £100 million.90
The defence departments also had to compete with political pressure
for extra Exchequer funds to help to deal with the housing problem that
had worsened during the war (when few houses were built), as well as to
pay for health services and education. Churchill, when arguing in 1924
for a reduction in the naval estimates, referred to the need to spend
money on social reform. Failure to do so, he warned, would lead to a
Socialist victory in the next election, and greater cuts in naval expen-
diture.91 There were also pressures to build roads and to provide other
social infrastructure. Between 1932/3 and 1938/9 civil expenditure of all
kinds that had to be paid for out of the chancellor™s budget increased by
£70 million, or 20 per cent.92
The other pressure on the chancellor was to reduce taxation. The
standard rate of income tax had increased from 1s 2d (5.8p) to 6s (30p)
in the pound between 1913/14 and 1918/19, and the top rate of income
tax plus surtax from 1s 8d (8.3p) to 10s 6d (52.5p). MPs, the press and
business interests agitated continually from 1919 for reductions. Lloyd
George responded by appointing the Geddes Committee in 1921 to
make recommendations on how expenditure could be reduced, and
most of the consequent cuts were at the expense of the defence
departments rather than social services.93
It is against this background that one can begin to understand the Ten
Year Rule and Churchill™s steps to strengthen it, ¬rst by having it
renewed in 1925, and then by persuading the Cabinet in 1928 to extend
it inde¬nitely, until such time as the Cabinet should decide to abrogate
it on the advice of the Foreign Of¬ce or the Chiefs of Staff. In March
1932 the Cabinet took that decision, in the light of Japanese aggression
in China, but the Chancellor, Neville Chamberlain, successfully resisted
any immediate rise in the defence estimates. In a Treasury forecasting
exercise in 1932 to see how the budget could be balanced in 1935 he
provided for an increase in the defence estimates from £103 million to
£115 million. In June 1934 he rejected a suggestion in a Cabinet
committee that some of the expenditure recommended by the DRC

Minutes of Inskip review, 25 Nov. 1937, T 161/855/S.48431/04, TNA.
Churchill to Baldwin, 15 Dec. 1924, in Gilbert, Winston S. Churchill, vol. V,
companion, part 1, pp. 303“7.
Peden, Treasury and British Public Policy, p. 296.
Daunton, Just Taxes, esp. p. 47; Andrew McDonald, ˜The Geddes Committee and the
formulation of public expenditure policy, 1921“1922™, Historical Journal, 32 (1989),
Retrenchment and rearmament 131

should be ¬nanced by borrowing; at that date the risk to ¬nancial
con¬dence, and therefore to recovery from the depression, seemed to be
greater than the risks of slow rearmament.94
However, on the initiative of the Permanent Secretary of the Treas-
ury, Fisher, it was decided in 1935 that the rearmament programme
being planned for the years 1936/7 to 1941/2 should be worked out on
the basis that a defence loan would be available. This change of heart
was brought about by intelligence reports about the scale of the German
government™s borrowing to pay for armaments. The Defence Loans Act
of 1937 authorised borrowing over the period 1937/8 to 1941/2, the
intention being to return to a balanced budget once the rearmament
programme was complete. The absence of the need to balance expen-
diture and taxation annually enabled the defence departments to spend
money more freely and to add to their programmes. The Treasury
worried that the defence forces being created would be too big to
maintain from taxation and tried to impose ¬nancial discipline by
˜rationing™ each department for the whole period down to spring 1942.
Financial allocations could only be made after priorities had been
reviewed on strategic grounds, and this task was given to Inskip, as
minister for co-ordination of defence, advised by senior of¬cials, the
most regular participants at meetings being Hankey; Sir Arthur
Robinson, chairman of the CID™s Supply Board; Sir Horace Wilson, the
chief industrial adviser; and Edward Bridges, the head of the Treasury
division dealing with defence expenditure. In the summer of 1937
the defence departments estimated future annual maintenance costs at
£255 million, on the basis of their approved programmes, and
£301 million on the basis of the expanded programmes that they
wanted. Treasury of¬cials believed that £220 million might be found for
defence from taxation from 1942/3, if there were no recession, and if
there were no sinking fund with which to reduce the national debt, but
the Cabinet was advised that it would be ˜most unwise to assume™ that
more than £150 million to £170 million would be available on the 1937
basis of taxation. The lower ¬gures were no doubt unduly pessimistic and
were intended to concentrate the minds of ministers on the need to
choose priorities or be prepared to raise taxes. In the event, although the
Inskip review curbed the defence departments™ tendency to add to their
programmes in the second half of 1937 and the ¬rst nine months of 1938,
further additions to the programmes, particularly between February and
June 1939, raised the estimated future maintenance cost of the defence
forces to £450 million, whereas 1939 tax rates (higher than in 1937)

Peden, British Rearmament, pp. 7, 67, 69.
132 Arms, economics and British strategy

were expected to produce only £250 million for defence. By June 1939
the Treasury was contemplating a rise from the current standard rate of
income tax of 5s 6d (27.5p) to a new, permanent, rate of 7s 6d (37.5p)
in the pound, higher than the maximum rate of 6s (30p) reached in the
First World War.95
Treasury arguments against unlimited defence expenditure were not
purely ¬nancial, although of¬cials did believe that repeated borrowing
would cause in¬‚ation, as it had done in the First World War. In 1937 they
used the concept of economic stability as a fourth arm of defence. On this
view, the real resources of the country were its manpower, productive
capacity, trade, and ability to raise loans. Britain had to import raw
materials and food, and nothing should be done to undermine industry™s
ability to export. The amount that the government could borrow
depended upon two factors: the savings of the community and con¬dence
in ¬nancial stability. If the government attempted to borrow more than
the money market was willing to lend, the Bank of England would have to
create credit and the result would be in¬‚ation. Higher domestic prices
would have an adverse effect on exports, savings would be reduced, and
¬nancial con¬dence would be weakened. In order to be able to wage the
long war that the Chiefs of Staff™s plans envisaged, Britain would have to
maintain its economic stability so that she would have greater staying
power than Germany.96 Chamberlain shared this line of thought,
although his concerns were not purely economic or strategic. On 25 April
1937, shortly before becoming prime minister, he noted that expenditure
on rearmament was causing prices to rise, and feared that in consequence
there might be ˜a series of crippling strikes ruining our programme, a
sharp steepening of costs due to wage increases, leading to the loss of our
export trade, a feverish and arti¬cial boom followed by a disastrous
slump, and ¬nally the defeat of the Government™.97
Economic stability was also necessary to maintain con¬dence in
sterling. Sterling was used to ¬nance international trade through
London, and overseas sterling balances were held there to facilitate
transactions. Sterling balances also formed all or part of the of¬cial
reserves of countries forming the sterling area: the dominions, except
Canada after 1931; India; British colonies, except Hong Kong; and

Ibid., pp. 40“2, 73“4, 88“9, 102“3.
˜Defence expenditure in future years: interim report by the Minister for Co-ordination
of Defence™, CP 316(37), 15 Dec. 1937, CAB 24/273, paras. 7“10. These paragraphs,
which the Chancellor of the Exchequer, Sir John Simon, called ˜a classic statement of
the elements that make up our strength for national defence™ (Cabinet conclusions,
16 Feb. 1938, CAB 23/92, TNA) are printed in Gibbs, Grand Strategy, vol. I, pp. 283“4.
Feiling, Neville Chamberlain, p. 292.
Retrenchment and rearmament 133

some semi-independent countries, like Egypt; as well as of some foreign
countries. During the Second World War overseas sterling balances held
in London were to rise from a normal pre-war level of about £500
million to £3,355 million, the difference representing the value of goods
and services obtained by Britain without payment other than a book
entry at the Bank of England. The international acceptability of sterling
was thus a valuable asset. However, foreigners would not hold more
sterling than was required for current transactions if sterling was
expected to depreciate against other currencies, as it tended to do in
1938“9 when Britain had an increasingly adverse balance of payments
on current account. Moreover, when sterling depreciated, the cost of
imports tended to rise, either directly as a result of goods priced in
foreign currencies becoming more expensive, or because sterling prices
abroad tended to rise. An increase in import prices was to be avoided
because 25 to 30 per cent of the price of armaments produced in British
factories represented the cost of imported raw materials. Foreign
exchange to ¬nance purchases outside the sterling area could be
obtained only by exporting goods and services; by drawing on gold and
foreign exchange reserves; by selling foreign securities owned by British
subjects, or by borrowing. There was, however, no immediate prospect
of loans from the American government, such as had sustained the
British war effort in 1917“18. Britain, like almost all other countries,
had ceased to make regular repayments of its war debt to the United
States during the depression of the early 1930s, and in 1934 Congress
included Britain in the provisions of the Johnson Act, which prohibited
new loans to any government that was in default.98
The Treasury™s arguments have been challenged by Christopher
Price, who claims that ˜had the will to rearm existed, British resources
were effectively limitless™. He believes that the system of imperial
preference agreed at the Ottawa conference in 1932, whereby Britain
and the dominions agreed to levy lower tariffs on trade between each
other than on trade with foreign countries, had provided the means to
escape the constraints of free-market conditions, but that instead Neville
Chamberlain preferred to appease the United States by agreeing in
November 1938 to a trade treaty that undermined the Ottawa agree-
ments. Britain, Price believes, could have matched Nazi Germany, with
its closed economy and intimidation of neighbours into joining its

Ian M. Drummond, The Floating Pound and the Sterling Area 1931“1939 (Cambridge
University Press, 1981); G. C. Peden, ˜A matter of timing: the economic background to
British foreign policy, 1937“1939™, History, 69 (1984), 15“28; L. S. Pressnell, External
Economic Policy since the War, vol. I: The Post-war Financial Settlement (London: HMSO,
1986), pp. 1“2, 413.
134 Arms, economics and British strategy

autarkic system, because members of the sterling area ˜were obliged to
accept sterling in payment for their commodities™.99 There are a number
of problems with this thesis. The dominions were independent countries
and they knew that the British market, large as it was, could not absorb all
of their exports of raw materials and food. They did not, therefore, wish
to cut themselves off from third markets by making the British Empire
autarkic. Nor were the dominions obliged to adhere to the sterling area “
Canada preferred to link its currency to that of its major trading partner,
the United States. Other countries in the Empire, particularly India,
the colonies, and areas of informal empire in the Middle East, had no
choice but to accept sterling in payment for goods and services, but these
countries, with the partial exception of India, were what would now be
called ˜underdeveloped™, and their exports were limited to food and raw
materials. It should also be noted that the sterling balances were not, as
Price assumes, an ˜unmitigated bene¬t™ to Britain.100 As will be discussed
in chapters 5 and 6, the sterling balances represented claims on British
exports. After the war they tended to make sterling vulnerable to spec-
ulative pressure, forcing successive governments to de¬‚ate by cutting
public expenditure, including defence expenditure.
Returning to 1938, one cannot assume that Chamberlain™s wish to
win favour in the United States was a sign of antipathy towards the
Ottawa agreements, of which he was one of the authors. Rather it almost
certainly re¬‚ected awareness that the Empire could not provide all the
goods that Britain would need in war, and that Britain would have to
look to the United States for supplies, especially of industrial goods, as
in the First World War. By 1938 Britain was already importing machine
tools and other engineering products required for rearmament from
outside the Empire, and experience in 1917“18 had shown that North
American sources of supply for all goods, including food and raw
materials, would be more economical in terms of shipping than more
distant Empire sources. Given the Johnson Act, imports from North
America would have to be paid for out of Britain™s gold or dollar
reserves, plus dollars earned from exports. Even cash purchases of
munitions would be problematic, given American neutrality legislation
in 1935 and 1937, which was designed to prohibit trade with, or credit
to, a belligerent. There were good reasons why Britain should seek ˜to
conciliate and please™ (to use Keynes™ words in 1916) both the United
States government and American public opinion in 1937“9.

Christopher Price, Britain, America and Rearmament in the 1930s: The Cost of Failure
(Basingstoke: Palgrave, 2001), esp. pp. xiii, 130“1.
Ibid., p. 185.
Retrenchment and rearmament 135

Price also believes that the depreciation of sterling and fall in the gold
and convertible currency reserves that occurred in 1938“9 could have
been halted at any moment by imposing exchange controls, instead of
allowing half of Britain™s war chest of gold and dollar reserves to be lost
in eighteen months before the outbreak of war. The sterling:dollar
exchange rate fell in 1938 from a high of $5.02 in March to a low of
$4.60 in September, and the Bank of England and Treasury used the
reserves to peg the rate at about $4.67, until losses became too great in
August 1939, when the rate fell to $4.10. Britain™s gold reserves fell from
£836 million on 31 March 1938 to £470 million on 22 August 1939.
Speculation against sterling over that period, Price argues, had nothing
to do with the state of Britain™s trade or economy. He demonstrates that
exchange controls were being considered, reluctantly, by economists in
the Cabinet Of¬ce as early as October 1937, and that opposition to
adopting controls on the German model was led by the Treasury. He
takes issue with the of¬cial historian of ¬nancial policy, R. S. Sayers,
who commented that nothing could have been done to prevent
foreigners withdrawing their money at the outbreak of war anyway.101
One problem with Price™s argument is that, given free movement of
sterling within the sterling area, exchange controls would have had to be
introduced by all countries within it, not just by Britain. Such co-
operation in the eighteen months before the war would not have been
impossible, but it would have taken time to arrange. Even after controls
had been introduced during the war, South Africa, the sterling area™s
major source of gold, was suspected of not co-operating fully. For
controls to have been effective all international transactions would have
had to be subject to government inspection, and postal censorship
would have had to be imposed. The economic cost would not have been
negligible; controls would have been an impediment to ¬nancing trade
outside the sterling area and would have had an adverse effect on
London™s position as an international banking centre, reducing invisible
exports in the form of income from ¬nancial services. At the end of July
1938 the Chancellor authorised the Bank of England to make pre-
parations for exchange controls, but these were only to be imposed after
the outbreak of war.102
Speculation against sterling after March 1938 was mainly in¬‚uenced
by ¬nancial markets™ reactions to political events. In December 1938

Ibid., pp. xiii“xiv, 85“6, 134“5, 150“1, 163“4: R. S. Sayers, Financial Policy, 1939“
1945 (London: HMSO, 1956), p. 229. For a balanced account, with data, see R. A. C.
Parker, ˜The pound sterling, the American Treasury and British preparations for war,
1938“1939™, English Historical Review, 98 (1983), 261“79.
Sayers, Financial Policy, pp. 227“32, 314.
136 Arms, economics and British strategy
Table 3.4. Balance of payments, 1930“8 (£ million at the end of calendar year)

(1) Visible (2) Invisible (3) Current (4) Capital (5) Change
balancea balanceb accountc accountd in reservese

À283 À8
1930 298 15 7
À322 À114 À34
80 f
1931 208
À216 À62
1932 154 91 29
À192 À18
1933 174 140 122
À220 À32
1934 188 42 10
1935 196 13 66 79
À263 À40
1936 223 251 211
À336 À57
1937 279 186 129
À285 À65 À203 À268
1938 220

Notes: a Export of goods minus imports of goods.
Net income from services and transfers of earnings.
Column 1 plus column 2.
All non-recurrent items, including new investment, payments on existing loans, net
short-term liabilities, and transfers between banks.
Column 3 plus column 4.
Includes £80 million assistance from foreign banks.
Source: Bank of England Quarterly Bulletin, 14 (1974), 47“52, at 49.

Sir Frederick Phillips, the of¬cial in charge of the Treasury™s ¬nance
divisions, believed that nine-tenths of the movement against sterling was
due to distrust of it as a currency in the event of war.103 However, Price
goes too far in claiming that speculation against sterling had nothing to
do with the state of Britain™s trade or economy. The adverse balance of
payments on current account in 1936 and 1937 (table 3.4) could be
attributed mainly to strong economic recovery from the depression, but
the existence of a growing balance of payments de¬cit at a time when
Britain was borrowing to ¬nance rearmament would have weakened
con¬dence in ¬nancial markets. Of the currency out¬‚ow of £268 million
in 1938 some £65 million, or about a quarter, was on current account
and therefore not the result of speculation or the defence of sterling.
With the bene¬t of hindsight, one can see that exchange controls
would have been worth imposing before September 1939, although at
what point after 31 March 1938 this should have been apparent is hard
to say. Sterling depreciation and exchange losses weakened Britain™s
capacity to wage a long war. Nevertheless, as we have seen, the growing
sterling problem did not prevent a rise in defence expenditure from
£262.1 million in 1937/8 to £382.5 million in 1938/9 or to an estimated

Parker, ˜Pound sterling™, 269.
Retrenchment and rearmament 137

£750 million in 1939/40, or a growth in the planned scale of the armed
forces from the programme as approved in mid-1937 (estimated to cost
£255 million to maintain) to what had been approved by mid-1939
(£450 million).

The defence industries
The pace of rearmament from 1936 was determined chie¬‚y by industrial
capacity and by the government™s policy of not interfering with normal
civil trade. Lord Weir, the industrialist and former secretary of state for
air who was consulted on the rearmament programme, advised ministers
in January 1936 that it could not be carried out in ¬ve years, as
recommended by the services, without affecting exports, unless semi-
war controls similar to those used in Germany and Russia were
imposed. Ministers did not dissent when he said that such controls
would be politically impossible in Britain.104 In September 1937 the
Secretary of State for Air, Lord Swinton, warned that a continuation of
the assumption that industry should not be interfered with would
postpone the completion of the aircraft programme for two years, from
1939 to 1941. Nevertheless, it was not until 22 March 1938, after the
Austrian Anschluss with Germany, that the Cabinet took the decision to
cancel the assumption, and it was another twelve months before War
Of¬ce orders enjoyed priority over normal trade.
With regard to the effects of the Ten Year Rule on the aircraft
industry, the of¬cial historian of British war production, M. M. Postan,
believed that government orders for aircraft prior to 1935 were insuf¬-
cient to sustain an airframe and aero-engine industry capable of
responding to the demands of rearmament for mass production of
modern designs.105 Peter Fearon has pointed to evidence that the Air
Ministry™s policy from the mid-1920s of allocating contracts so as to
maintain a ˜ring™ of as many as eighteen ¬rms in business prevented some
¬rms gaining production experience.106 However, David Edgerton
and Sebastian Ritchie have shown that the British aircraft industry was

Defence Policy and Requirements (Defence Requirements) Committee (DPR(DR)C)
minutes, 13 Jan. 1936, CAB 16/123, TNA.
M. M. Postan, British War Production (London: HMSO, 1952), p. 5.
Peter Fearon, ˜Formative years of the British aircraft industry™, Business History Review,
43 (1969), 476“95; Peter Fearon, ˜The British airframe industry and the state, 1918“
35™, Economic History Review, 27 (1974), 236“51; Peter Fearon, ˜The vicissitudes of a
British aircraft company: Handley Page Ltd between the wars™, Business History,
20 (1978), 63“86; Peter Fearon, ˜Aircraft manufacturing™, in N. K. Buxton and
D. H. Aldcroft (eds.), British Industry between the Wars: Instability and Industrial
Development, 1919“39 (London: Scolar Press, 1979), pp. 216“40.
138 Arms, economics and British strategy
Table 3.5. Military aircraft production in Britain, France, Germany, Japan, the USA and the
USSR, 1933“40

(1) Britain (2) France (3) Germany (4) Japan (5) USA (6) USSR

1933 633 n/a 368 766 466 2,595
1934 n/a 1,968 688 437 2,595
1935 785 3,183 952 459 3,578
1936 1,830 890 5,112 1,181 1,141 3,578
1937 2,218 743 5,606 1,511 949 3,578
1938 2,827 1,382 5,235 3,201 1,800 7,500
1939 7,940 3,163 8,295 4,467 2,195 10,383
1940 15,049 n/a 10,247 4,768 12,804 10,565

Notes: a In addition 298 were exported.
In addition 453 were exported.
Sources: Col. 1: Sebastian Ritchie, Industry and Air Power: The Expansion of British Aircraft
Production, 1935“1941 (London: Frank Cass, 1997), pp. 9, 90; cols. 2“6: R. J. Overy, The
Air War 1939“1945 (London: Europa Publications, 1980), pp. 21, 150.

one of the strongest in the world and did not depend solely on domestic
orders. Britain was the largest exporter of military aircraft from the late
1920s to the mid-1930s, with exports accounting for 26.9 per cent of all
British production, including civil orders, in 1934 and 25.1 per cent in
1935. Moreover, there was a trend towards consolidation: Vickers
purchased Supermarine in 1928; and by 1935 the Hawker Siddeley
group included the Armstrong Whitworth, A. V. Roe, Gloster and
Hawker airframe companies and the Armstrong Siddeley aero-engine
company. Successful design teams, notably Fairey in the 1920s and
Hawker in the 1930s, could make aircraft production a pro¬table
business. Similarly, aero-engine production came to be dominated by
Bristol and Rolls-Royce, who successfully specialised in the develop-
ment of air-cooled radial and water-cooled in-line engines respectively.
Prior to 1934 the British aircraft industry employed about twice as many
workers as the German. Moreover, the German aircraft industry was not
notably more concentrated than the British in the 1930s, with eleven
¬rms of various sizes producing airframes (compared with fourteen in
Britain) and ¬ve producing aero-engines (the same as in Britain).107
British aircraft production fell behind that of Germany in the mid-1930s,
but almost caught up in 1939 and was nearly 50 per cent higher than
Germany™s in 1940 (see table 3.5). The reasons for any shortcomings

Edgerton, England and the Aeroplane, pp. 24“8; Edgerton, Warfare State, pp. 42“4;
Sebastian Ritchie, Industry and Air Power: The Expansion of British Aircraft Production,
1935“1941 (London: Frank Cass, 1997), pp. 9“19.
Retrenchment and rearmament 139

of the British aircraft industry are to be found in the years 1934 to 1938
rather than earlier. It was not until September 1936 that Swinton felt he
could say that as important a ¬rm as Vickers was ˜at last fully alive to the
necessity of straining every nerve to carry out their orders™ under the
rearmament programme.108
One problem was that, despite increasingly generous contracts, it was
dif¬cult to persuade British ¬rms to expand plant to the full extent
required, given their experience of being left with surplus capacity after
the First World War, when some ¬rms, including one of the largest,
Sopwith, had gone bankrupt, and given that a reduction in orders could
be expected in 1942, when the rearmament programme was due to be
completed.109 A related problem was the need to prepare ¬rms outside
the aircraft industry to take part in the greatly increased production that
would be required if war broke out. These problems were met by
creating a ˜shadow™ industry for aircraft production. Plans for creating
additional capacity for aircraft production had been drawn up in 1929
and were implemented from 1936. At ¬rst the shadow industry was
intended to be made up of factories managed by automobile and other
engineering ¬rms, which would be given ˜educational™ orders in peace-
time but whose main effort would be made in war-time. However, in
1938 the government decided to expand productive capacity by paying
for the construction of factories that would be managed by established
aircraft ¬rms for an agency fee, and it was this expansion that made the
greatest and most immediate contribution to productive capacity. There
was also a rapid increase in sub-contracting from the spring of 1938,
when the aircraft ¬rms were told by the Air Ministry that they should
aim to sub-contract at least 35 per cent of aircraft construction, whereas
10 per cent had been more normal earlier in the rearmament pro-
gramme. As ¬rms gained experience, labour productivity increased.110
Down to the spring of 1938 the Air Ministry had avoided placing
orders for obsolescent aircraft merely to boost production ¬gures. On
27 April 1938, however, the Cabinet, under parliamentary pressure to
match German aircraft production, authorised the Air Ministry to
accept as many aircraft as it could from the British aircraft industry, up
to a maximum of 12,000 machines over the next two years. The
Treasury readily sanctioned expenditure on plant required for the
programme. As a result of these measures the gap between British and

Secretary of State™s progress meeting, 15 Sep. 1936, Air Ministry records, series 6, ¬le
26 (AIR 6/26), TNA.
Secretary of State™s progress meeting, 17 Sep. 1935, AIR 6/23, TNA.
William Hornby, Factories and Plant (London: HMSO, 1958), pp. 218“26; Ritchie,
Industry and Air Power, pp. 57“61, 91“105.
140 Arms, economics and British strategy

German production rapidly narrowed. The performance of Britain™s
aircraft industry compared very favourably with that of France, where
nationalisation was associated with a fall in production in 1937, and
where total production in the years 1936 to 1939 inclusive, 6,178, was
only 41.7 per cent of the British total (see table 3.5).
The conventional picture of Britain™s naval armaments industry is that
it contracted considerably after the First World War and by 1935 was
not capable of producing as many warships a year as it had done before
1914.111 However, it still had the bene¬t of supplying the largest navy in
the world. Britain may have had to concede parity to the United States
Navy under the Washington and London naval treaties, but, as table 3.6
shows, Britain continued to order more warships than other naval
powers both before and after rearmament began in earnest in 1936.
Consequently, as Edgerton has shown, Britain had the world™s leading
naval-industrial complex. The industrial capacity involved extended far
beyond the shipbuilding industry as conventionally de¬ned: for exam-
ple, 40 per cent of the price of a battleship represented the cost of guns,
gun-mountings and armour plate. In the 1920s and early 1930s warship
tonnage ordered averaged about half of the average for the decade
1900“10. Some ¬rms did not prosper “ notably Beardmore, which
ceased to build warships (although it continued to produce armour and
guns) after 1930. Others did much better. Vickers, half of whose turn-
over was in armaments, was the country™s third largest manufacturing
employer in 1935. From 1925 the Admiralty provided subsidies, direct
or through the pricing of contracts, to ensure that specialised industrial
capacity did not disappear: for example, for armour and shells. There
was spare capacity waiting to be reactivated in 1935 and, while there
were procurement bottlenecks in the supply of guns, gun-mountings,
armour plate and electrical equipment, the Admiralty had to spend only
a modest £12 million on new plant for private contractors, Admiralty
factories and royal dockyards between April 1936 and April 1939,
compared with commitments totalling £40 million made by the Air
Ministry in the slightly longer period April 1936 to July 1939.112
By contrast, the War Of¬ce was very much constrained by lack of
industrial capacity. Of the 250 national factories created by the Ministry
of Munitions in the First World War, only three were retained through
the inter-war period, and these were held in reserve and not rehabili-
tated until 1936“7. Munitions production down to 1936 was shared
between the three historic royal ordnance factories at Woolwich,

Gordon, British Seapower is the fullest exposition of this point of view.
Edgerton, Warfare State, pp. 33“41; Hornby, Factories and Plant, pp. 202“3.
Retrenchment and rearmament 141

Table 3.6. Major warships launched (or conversions to aircraft carriers begun), 1922“35 and
1936“40 a

Britain France Germany Italy Japan USA

(3 ¼ 1)b
Capital ships 2 1 0 0 0
3c 3d
Aircraft carriers 1 0 0 5
Cruisers 18 6 17 23 21
Destroyers 62 56 5 40 59 21
13 f
Submarines 34 77 12 54 53
Capital ships 5 3 4 4 2 2
1g 1h
Aircraft carriers 5 0 6 4
Cruisers 28 1 5 3 5 11
Destroyers 69 10 25 16 32 73
Submarines 39 6 140 62 27 33

Notes: a Excluding monitors and minelayers, and all surface ships of under 1,000 tons.
Armoured ships, popularly known as pocket battleships, each with a tonnage of about
one-third of the maximum for a capital ship under the Washington Treaty.
Not including two wartime conversions or HMS Hermes, the world™s ¬rst ship designed
from the outset as a carrier, which was launched in 1919.
Excluding Langley, converted from a ¬‚eet collier in 1920“1.
A further ¬ve were built for the Royal Australian Navy.
Launched 1938; construction suspended 1940.
Conversion of passenger liner begun in October 1940.
Source: Roger Chesneau (ed.), Conway™s All the World™s Fighting Ships 1922“1946 (London:
Conway Maritime Press, 1980).

Waltham and En¬eld, and a much reduced private sector. There was
adequate capacity for manufacturing small arms and ammunition for the
army™s rearmament programme of 1936, but not for guns, shells or
tanks. Ten new royal ordnance factories were approved between 1936
and 1939, but this capacity was based on the size of army planned in
1936, not the army as it was expanded after March 1939. Re¬‚ecting the
strategic priority given to air war, much of the new capacity for manu-
facturing guns was devoted to anti-aircraft requirements rather than
those of the ¬eld army.113 The Air Ministry was allowed by the Treasury
to pre-empt War Of¬ce requirements for machine tools in 1936 and, as
part of the policy of non-interference with normal trade, the Treasury
tried to have War Of¬ce orders postponed until they could be ful¬lled by

Hornby, Factories and Plant, pp. 83, 89, 148“9; Postan, British War Production, pp. 7“8, 33.
142 Arms, economics and British strategy

the new royal ordnance factories, rather than being used to bring new
¬rms into munitions production. The Morris car company and the
London, Midland and Scottish Railway were asked to co-operate in
the design and production of tanks, but in April 1939 Sir Harold Brown,
the director-general of munitions production, reported to the CID that
˜until recently™ tank production had not been given a very high degree of
priority as the work con¬‚icted with locomotive and railway wagon
Regarding the rearmament programme as a whole, Weir identi¬ed the
supply of skilled labour as the most important bottleneck, and advised in
January 1936 that steps should be taken to secure ef¬ciency in its use.115
As in the First World War, changes in industrial practices “ dilution “
could overcome shortages of skilled labour, but trade unions were
deeply suspicious that employers were trying to turn the situation to
their own advantage, and feared that unemployment would follow the
completion of the rearmament programme. The advice of the Ministry
of Labour was that direct government contact with the trade unions
should be avoided, as consultation would encourage the unions to
demand a high price as regards conditions and wages in return for co-
operation. The Ministry recommended using sub-contracting to take
work to areas where there were still reserves of unemployed labour, and
relying upon upgrading of individual skilled workers to better-paid
positions supervising the work of semi-skilled and unskilled labour. The
political atmosphere changed after the Anschluss and contacts between
the government and trade union leaders began the same month (March
1938). Even so, it was not until 28 August 1939 that the Amalgamated
Engineering Union and the Engineering Employers Federation at last
signed an agreement on dilution. As R. A. C. Parker argued, there were
real economic limits on rearmament.116
As in the First World War, private capital was attracted to armaments
production by the prospect of pro¬ts, a sensitive matter for trade
unionists and the wider political public. Neville Chamberlain tried to


. 5
( 14)