How Imperial was the Committee of Imperial Defence?1

Peter Catterall
Institute of Contemporary British History
London, England

In 1942 the Deputy Prime Minister, Clement Attlee, commented, �Have we not found in dealing with the Dominions that the more we avoid precise definitions, the better� (Attlee, 1942). Indeed, the 1931 Statute of Westminster, which at the time was described rather grandiosely by some as a written constitution for the Commonwealth (Elliott, 1930), was instanced by Attlee as the type of more precise definition which ought to be avoided. Instead, relations at the core decision-making level of the Cabinet, which is what Attlee was discussing, were characterised more by pragmatism and flexibility. Arrangements on consultation or entitlement to sit on key bodies at this level were largely unformulated, only becoming an issue when they were, as happened from time to time, the subject of conflict between Britain and one or more Dominions. There were only a few exceptions to this rule. It was, for instance, recognised by British officials that Dominion Prime Ministers - though not their Cabinet colleagues (Churchill, 1944) - were de facto members of the War Cabinet during both World Wars, if in London and except for domestic purposes, just as they had been de facto members of the Committee of Imperial Defence [CID] in peacetime (Rowan, 1945). This formal rule, however, tells us little about how imperial relations operated at this level during most of the time when such special circumstances did not obtain.

The informality of imperial relations at the level of the Cabinet and the CID does not indicate that they were insignificant. Britain in the first half of the twentieth century was at the centre of a far-flung maritime empire with numerous strategic problems and responsibilities. And during this period it not only had to confront many minor conflicts on the periphery, it also had to cope with two World Wars which required the mobilisation, as far as possible, of the entire resources of the empire for victory. Defence requirements were therefore not only a fundamental instrument of the institutional changes in British government effected in these years (Catterall & Brady, forthcoming). They were also an important common concern shared both by the metropolitan power and by the self-governing territories of the empire. Indeed, the century opened with the first significant example of these Dominions sharing in Britain�s warfare burden, in South Africa.

This conflict was to have a major impact upon British defence arrangements. Already, in the late nineteenth century an increasingly overstretched and internationally threatened imperial power was making tentative moves to improve the co-ordination of its defence administration and planning. The recommendations of the 1890 Hartington Commission that �a naval and military council, which should probably be presided over by the Prime Minister� (Johnson, 1960: 32) was, however, only inadequately followed. The Cabinet Defence Committee established in 1895 had no clear relationship with either the Cabinet or the service ministries, let alone the empire, and proved, in the words of Balfour�s private secretary, Jack Sandars, �an absolutely useless body� (Mackay, 1985: 66-8). In the aftermath of the Boer War, however, Balfour, who became Prime Minister in 1902, moved to replace it with a rather more effective body. Following memoranda from the two service ministers, Broderick and Selborne, the Defence Committee was reconstituted in November 1902. In 1903 the Prime Minister himself took over the chairmanship and it was renamed the Committee of Imperial Defence. The following year, as a result of the report of the Esher Committee set up in the aftermath of the Boer War, it acquired a secretariat which was to become the basis for the future Cabinet Office.

At the time an even more radical remodelling of the British defence organisation was being canvassed. From the 1890s figures such as Sir Charles Dilke had advocated the creation of a Ministry of Defence. This option, however, was rejected by Balfour on a number of grounds. Firstly, recent experience, not least in the sometimes chaotically fractious circumstances of the last Salisbury government (Otte, 1996), had suggested that a Minister of Defence would not have enough authority to co-ordinate the services. Instead, therefore, this role had to be discharged by the Prime Minister. As the Esher Report put it;

As a result the CID, in principle at least, consisted of the Prime Minister and whosoever he choose to appoint to assist him.

In practice, CID came to acquire a regular membership of certain of the great officeholders of state (Hankey, 1937), many of whom were more regular in their attendance than the Prime Minister himself. Nevertheless, this founding principle reflected the second reason why Balfour and others rejected the idea of a Ministry of Defence; they sought an organisation which would be able flexibly to co-ordinate defence policy. This was not just a matter of co-ordinating policy across the service ministries, but across the whole network of departments which, as the Great War was vividly to demonstrate, were all involved in defence considerations. CID and the position of the Prime Minister within it were thus both means of ensuring effective co-ordination throughout Whitehall. And the network of sub-committees which rapidly developed could bring the expertise of varying combinations of departments to bear upon the issues they addressed. 2

But the articulators of the system were at pains to point out that this did not derogate from the responsibility of ministers for the actions of their departments. To ensure that �the final power of decision by the Cabinet, as the body responsible to Parliament, is preserved� (Hankey, 1936: 25), CID formally only had advisory powers, even if, in the vast majority of cases, its recommendations were adopted without requiring Cabinet approval (Ismay, 1939: 5). As its functions largely concerned planning and preparation for the event of war, this advisory character was entirely appropriate until the actual outbreak of war, when the need instead for decision-making structures called for different instruments. Accordingly, during the First World War CID ceased to meet early in 1915 and it was immediately wound up on the outbreak of World War Two. CID was thus effectively an advisory peacetime defence planning system in which formal authority remained with departmental ministers. This ensured not only flexibility, but also its acceptability to ministers.

If these were requirements which applied to the relationship between CID and British departments of state they applied a fortiori to the relationship between CID and the empire. This furnished the third and final reason why Balfour preferred the flexibility of CID to a Ministry of Defence. Some kind of defence relationship was desirable with the self-governing territories of the empire which were increasingly developing their own defence capacity. CID could function as the one institution, apart from the Crown, which spanned the empire. Indeed, some kind of co-ordination was already developing under the Colonial Defence Committee set up in 1885, which after 1904 became a sub-committee of CID, though this was essentially confined to local defence arrangements (Johnson, 1960: 18-21). But there were also problems, for both Britain and the Dominions, with moves to more formal arrangements.

This is not to say such attempts were not made, particularly on the British side, from the first Colonial Conference in 1887. This conference took place against the background of an apparently growing challenge from Russia, and war, or the threat of war, was to continue to play an important part in moves towards imperial co-ordination. For instance, the need to secure imperial support in the face of the Anglo-German rivalry of the Edwardian years culminated in the admission of the Dominions as full members of the CID at the 1911 Colonial Conference. Even so, it was not until after three years of war, in 1917, that an Imperial War Cabinet [IWC] including Dominion ministers was formed. This, moreover, was little more than a formal recognition of the imperial plenum which had already been established by the attendance in London of Dominion Prime Ministers at successive Colonial Conferences. It grew, as Lloyd George pointed out, �not by design, but out of the necessities of the war�. In particular, as the War Cabinet Report for 1917 explained, �some method had to be found of informing the Overseas Governments of the political and military situation�. Featuring the Prime Ministers of New Zealand, Canada, Newfoundland and Australia (in 1918) 3, the South African Minister of Defence, J C Smuts, and representatives of Princely and British India, it met in concert with the British War Cabinet in four sessions in 1917 and 1918 to discuss general matters relating to the war and, finally, the preparations for the peace settlement.

There are four points to be made about the IWC. Firstly, it was a pragmatic response to the situation, rather than a deliberate act of policy. Neither the British nor the Dominions set out to create it, and some of its actions, such as the decision in 1917 that India should in future be fully represented at imperial conferences reflected the military contributions and exigencies of the moment. Secondly, it was, as Lloyd George put it, �extremely elastic�. It did not formally commit its members, indeed, in constitutional theory it could not. As Sir Robert Borden, the Canadian Prime Minister, pointed out, each of the Dominions represented preserved �unimpaired its perfect autonomy, its self-government, and the responsibility of its Ministers to their own electorate�. Accordingly, it could only be consultative. Thirdly, however, it was possible to see the IWC as the basis of more permanent institutions. Borden, for instance, in 1917 argued that post-war adjustments to imperial structures �should provide effective arrangements for continuous consultation in all important matters of common Imperial concern and for such concerted actions, founded on consultation, as the several governments may determine�. This idea of giving some kind of permanence to the IWC was enshrined both in the idea of annual IWCs Lloyd George floated in May 1917 (War Cabinet, 1917) and in the resolution the IWC passed on 30 July 1918 that Dominion Prime Ministers might nominate one of their Cabinet ministers as resident in or visitor to London to represent them between its plenary sessions (War Cabinet, 1918).

This leads on to the fourth point about the IWC, that it was essentially, as in the pre-war conferences, a meeting of heads of autonomous governments. It thus reflected in embryo the relationships that Balfour was to encapsulate in his description at the 1926 Imperial Conference of Britain and the Dominions as �autonomous communities within the British Empire, equal in status, in no way subordinate one to another in any aspect of their domestic or external affairs, though united by a common allegiance to the Crown� (Macintyre, 1986: 206), a formulation that was to be formally enshrined in the 1931 Statute of Westminster. The Dominions, by these developments, became responsible for their own foreign policy and defence. The distinguished Australian constitutional authority, A B Keith, was later to argue that as a result it was impossible to recreate the IWC during World War Two (Keith, 1942). Certainly the IWC was not recreated in the same form, though it might be argued that the dissimilarities between the two conflicts accounts for this, rather than constitutional changes. For instance, imperial consultation was of greater relative importance in the First World War, and the IWC pre-dated the creation of the Supreme War Council of the Allies, which first met in November 1917. In contrast, during the Second World War, Anglo-American relations were of supreme importance, and certainly recognised as such by the Canadians (Minutes, 1943).

However, nor was the IWC recreated during the inter-war years, despite the resolutions of 1917-18. In part this reflected the political impossibility in peacetime of the kind of delegation envisaged in the July 1918 resolution. Such ministerial delegates either risked becoming captives of the British government, or plenipotentiaries only in name, divorced from their governments back home. As far as the British were concerned, meanwhile, these delegates might make demands on the metropolitan power which did not accord with its perceived interests. The experiment of creating a sort of standing imperial council out of the IWC was thus stillborn. In the inter-war years there was instead a reversion to the pre-war system of imperial liaison embodied in occasional conferences and in the CID network, which began to re-emerge in early 1919, although the CID itself did not meet until 1920.

The re-emergent CID recreated a pattern in terms of imperial relations which had already become established in the Edwardian period. However, there were some qualitative developments as a result of the First World War. Some of these were positive. For instance, the 1918 Imperial Conference had emphasised the importance of improving imperial communications (War Cabinet, 1918: 12). The result was the establishment of an imperial communications sub-committee of the CID in 1919, its importance being reflected in the fact that it was one of the few CID sub-committees to have a ministerial chairman.

Whether the legacy of the war actually meant communication improved in terms of the transaction of CID business was, however, another matter. Before 1914 CID had begun to act as a defence planning agency for the empire, providing occasional advice for Dominions, a function that it to some extent continued to perform in the 1920s (Johnson, 1960: 88, 213). But communication was routed through ministers, indeed the Dominions had refused in 1911 to allow representation at CID through their High Commissioners, though Borden had decided to change this in 1914 (Johnson, 1960: 124). This avoided them being drawn into the net of British planning and asserted political control over their handling of imperial relations. Indeed, as a result, in the aftermath of the First World War the British were moved to complain that the CID documents they sent to the Dominions did not seem to get beyond the Prime Ministers� offices to the departments to which they were addressed (Hankey, 1927).

It was to avoid such political problems that Balfour had originally emphasised that CID was advisory. This reassurance, however, had to be reiterated in the 1920s. In April 1928 British attempts to persuade Dominion governments to allow their High Commissioners to act as representatives on CID, not least through a tour of Dominion capitals by the Dominions Secretary, Leo Amery, relied heavily upon the assurance that CID, as an advisory body, could not commit them to specific actions. This assurance was welcomed by the Australians and South Africans, and they and the New Zealanders now agreed to representation when required through the High Commissioners. No attempt seems to have been made, in the face of obvious political difficulties, to persuade the Irish, even before the advent to power of de Valera and the conflict over the oath of allegiance which began in 1932. 4 And British attempts to win Canadian participation meanwhile failed. They found it impossible to persuade Mackenzie King of the advisory nature of CID (Hankey, 1929), just as they had had similar difficulties with Borden�s predecessor, Sir Wilfrid Laurier.

This attitude on the part of the Canadians was to prove a matter of considerable frustration for the British, not least as the next war drew nearer. For instance, liaison over the maintenance of food supplies was clearly a matter of cardinal importance to them, given the need to import across the soon to be submarine-infested waters of the Atlantic. However, they found the Canadians very reluctant to discuss this (Campbell, 1939).

The British, bearing the brunt of imperial defence throughout this period and at the sharp end of the repeated threat from Germany, looked for support from the Dominions. Hence, for instance, the special Colonial Conference on imperial defence at the time of the Dreadnought race in 1909. Rightly, they saw themselves as the principals in the imperial defence effort. They therefore sought to pursue what they saw as the necessary requirements. This was not unsuccessful. In advance of the First World War co-ordination machinery, involving the compilation of preparatory local War Books, was set up in Britain and the Dominions. From the establishment of the Co-ordination sub-committee of the CID in 1911 the relevant sections of the British War Book it created were sent out to the Dominions. And on the outbreak of war the integration of imperial forces proceeded with a smoothness which sometimes eluded the wartime co-ordination of the Royal Navy and the British Army.

This was, however, a circumscribed success, not least in the clear limitations, even before the First World War, to Dominion willingness to accept the British lead. In particular, the Dominions did not wish entanglement in European conflicts. Therefore what they sought within the CID structure was not integration, which would necessarily place them in a subordinate position, but information. In 1911, for instance, the Australians demanded that they be consulted over European developments, a demand which led to Asquith�s concession of a right of representation to the Dominions at CID (Macintyre, 1986: 140; Johnson, 1960: 110). This need for information and consultation was reinforced in the aftermath of the Great War, not least during the Chanak crisis of 1922, in which the British essentially assumed Australian support. To ensure Australia was subsequently better informed on imperial policy Richard Casey was appointed as Australian liaison officer in the Cabinet Office in London, in which post he served from 1924-31.

For Australia and New Zealand during the inter-war years there was also another reason for their need for information and consultation. Conscious of their own vulnerability, especially in the face of the rising threat of Japan, they continued to look to Britain for protection. As the Australian Premier, Sir Robert Menzies, commented early in the Second World War, �What Great Britain calls the Far East is to us the near north� (Macintyre, 1986: 327). Accordingly, the Australians in particular had reasons for seeking leverage on the British which were not shared to the same extent by other Dominions. This was reflected in their attitude to the War Cabinet during the Second World War.

After the various experiments which culminated in the Lloyd George War Cabinet during the First World War, considerable thought was given, particularly from 1938 onwards, on how best to establish a War Cabinet on the outbreak of hostilities. For several commentators the role of the Dominions in the previous war meant that they had to be represented, even if only in the person of the Dominions Secretary, from the start of the Second World War (Yates, 1939; Laski, 1939). The Secretary of CID, H L Ismay, however early on took the view that the means of association with the Dominions was a question which could be settled later (Ismay, 1938). In the event the Dominions Secretary did not become a member of the War Cabinet, although he usually was able to attend when matters within his purview were under consideration.

Mackenzie King pronounced himself satisfied with these arrangements in the Canadian House of Commons in January 1942, even though the Canadians were later to complain at a meeting with British ministers in 1943 of occasional lack of consultation, especially regarding Anglo-American negotiations (Minutes, 1943). By late 1941, however, in the face of the growing Japanese threat, the Australians were definitely not satisfied. In August 1941 Menzies called for direct Dominion representation in the London War Cabinet in the face of the situation in the Far East, offering to go himself to London, where he had already spent a large part of the year. This offer, however, essentially reflected his domestic political difficulties in the aftermath of the indecisive 1940 Australian elections. Although the Australian public were supportive, the British government were not. It was pointed out that there was no precedent for such a move. General Smuts, who had attended the War Cabinet in 1917-19, had done so in a personal capacity, not as the representative of the Union of South Africa (Cranborne, 1941). Menzies� overtures were therefore rejected, and he resigned the premiership shortly afterwards.

The government changed to Labour in October 1941, but Australia�s strategic problems continued, and therefore so did the anxiety for representation at War Cabinet level. This became even more acute with the Japanese landings in Malaya in December 1941. As the Japanese troops advanced towards the key strategic fortress of Singapore, which was to fall the following month, the Dominions Office on 27 January 1942 cabled British willingness to accept Dominion representation on the War Cabinet (Dominions Office, 1942). The Australians and New Zealanders immediately took up the opportunity to be represented by their High Commissioners. Later in the year India, by now also threatened by the Japanese, was also accorded representation. It, indeed, had two representatives as in the Great War, one of Princely and one of British India. Churchill�s attempts to restrict this to one representative were overruled by the India Office on grounds of political impossibility, prompting the Prime Minister, ever mindful of the ideal of the War Cabinet as a small, decision-making elite, to growl �We shall have to take the Albert Hall for our War Cabinet meetings� (Churchill, 1942a).

The need for representation was greatest for those Dominions (and India) which felt most threatened. And although ties of sentiment were undoubtedly there for both Australia and New Zealand, their desire for representation at the highest level in London was tied to their particular requirements; witness their similar interest in the Pacific War Council also set up in 1942. There was thus a certain instrumentality in their approach to the War Cabinet. But the same was true of the British. They made it clear that the High Commissioners and the Indian delegates were no more than representatives, and in no sense were they members of the War Cabinet. They also clearly regarded this as an anomalous situation brought about by the Far East emergency, and one they had to bear, particularly in their relations with S M Bruce, the Australian High Commissioner, with barely concealed irritation. Therefore the suggestion of Australian Premier John Curtin at the May 1944 Imperial conference that representation on the War Cabinet be replaced by a monthly meeting for High Commissioners was seized on with alacrity by the British (Bridges, 1944). Given the contemporary difficulties of British policy in India, they were no less keen to see the back of the Indian delegates, who also ceased to attend in 1944.

This attitude seems to contrast with the enthusiasm with which the then Secretary of the Cabinet and the CID, Sir Maurice Hankey, welcomed Casey in 1924 (Roskill, 1972: 401). Even so, although he seems to have had access to all the CID papers, Casey did not get to see them all as a matter of routine. There was a limit to how much the British would share the secrets of imperial defence even with someone on the spot like Casey. However, even more, there was a limit to how much the Dominions were themselves prepared to contribute to imperial defence. What they sought was information, not the sharing of the defence burden. In the run-up to the Second World War, for instance, despite their continuing anxieties about Singapore, Australia was spending no more than 1 per cent of national income on defence and made clear, at the 1937 Imperial Conference, their reluctance to support the pursuit of a tougher policy by Britain in Europe. Indeed, the only Dominion which was prepared to talk tough in 1937 was New Zealand which, ironically, spent even less proportionately on defence than Australia (Crozier, 1997: 136-7).

Britain, however, formally renounced any efforts to cajole the Dominions over defence at the 1926 Conference. This policy Casey regarded as misguided (Roskill, 1972: 429), but it was also politically realistic. The British were well aware of the sensitivities and difficult political situation in, for instance, South Africa, having some anxiety as to whether this strategically important Dominion might prefer neutrality on the outbreak of war (Roskill, 1974: 396). But even in Australia there were palpable political problems, which continued after the outbreak of war. The British High Commissioner, for instance, noted in 1941 that Curtin as Labour leader could not support an all-party coalition similar to that in Britain, given the strength of neutralism in his ranks (Whiskard, 1941). Any British attempts to steer policy in such delicate political situations ran a high risk of backfiring. Instead, such efforts as were made were deliberately informal, such as Hankey�s tour of Dominion capitals in 1934 in the light of the proceedings of the high-powered official CID sub-committee on defence requirements which he chaired, and which was to lay the basis for British rearmament from the following year. The fact that Hankey was an official meant that the visits could be seen as a sharing of information and a means of prompting local planning, rather than the bringing to bear of imperial pressure (Johnson, 1960: 218).

In practice, in any case, the sharing of information by the British could be highly selective. Which papers were to be sent to Dominion governments were picked out by the CID secretariat on a need-to-know basis (Hankey, 1922: 16). Nor were the Dominions, even after the decision of 1928, any more heavily involved in the actual proceedings of CID. As war threatened in the 1930s the frequency of meetings of CID and its sub-committees and the numbers of different individuals attending steadily rose. Thus in 1934/35 there were five meetings of the CID itself and 198 meetings in all, involving 548 persons. Of these just nine came from the Dominions. By 1938 the number of meetings had risen to 38 and 409 respectively. Dominion representation, however, fell from the peak of 22 persons in 1937 to a mere 14. This was out of a total of 814 (in 1937) and 876 (in 1938) attending CID meetings. Not only was Dominion representation minuscule, it was also far outstripped by the number of persons attending CID meetings simply as non-government experts. 5

Formal Dominion representation on CID committees was similarly limited. High Commissioners, after all, could not commit their governments. Accordingly, representation and reporting back was usually through the medium of the Dominions Office, which had been set up in 1925. There were few committees which included direct Dominion representation amongst their formal membership. Essentially these were confined to those dealing with supply matters. A number of these committees had Australian, New Zealand and Indian associate members.6 Otherwise, the High Commissioners or their staff �were only to be invited if and when the Prime Minister of the United Kingdom thought advisable�. If, that was, there was something on the agenda which the Prime Minister considered within their field of interest (Hankey, 1930: 6). Generally this was taken to mean matters of regional interest, on which the Dominions might have a useful contribution to make. The result could, especially in times of tension, be that the High Commissioners complained about the lack of consultation and a certain feeling that they were only invited when unimportant matters were being discussed. This was not without justification, as the British Minister for the Co-ordination of Defence recognised in August 1939. However, as he also went on to point out, the British had to exclude the High Commissioners from the discussion of certain secret weapons, or when such matters as delays in the despatch of the fleet to the Far East - which probably would not go down well with the Australians - came up on the agenda (Chatfield, 1939a).

Just as the Dominions, then, had political difficulties over their degree of involvement in the CID system, so did the British over the degree to which they wished to involve them. For the latter as much as the Dominions themselves, the main point of Dominion involvement in CID was to share information. More formal arrangements not only posed political problems for the Dominions, but for the British as well. Dominions could, for instance, put pressure on concerning their own special interests. The British did not enjoy feeling that they were being made beholden to the Dominions in this way, especially when a reciprocal commitment was not always readily apparent. Churchill, for instance, complained in 1943 about having to put up with Bruce at War Cabinet meetings, despite Australian troop withdrawals (Churchill, 1943). And the search for information by the Dominions could feel to the British like intrusion into what were properly their own affairs. As Churchill exasperately minuted, �We do not sit in the Australian Cabinet� (Churchill, 1942b).

In peacetime there was less of a problem. The Dominions felt less urgency to acquire information about imperial policy. The British were therefore under less pressure regarding the sharing of that information. They could plan through the CID system, involving the Dominions as much or more usually as little as they considered necessary. Important issues which had major domestic implications, usually finance, might even be taken out of the CID system altogether. Examples were the committees on expenditure or the naval programme in the 1920s, which were instead set up as Cabinet committees centrally concerned with finance. CID, as an advisory system, was not particularly well-suited to tackle such matters, which is also why decisions on rearmament in the 1930s tended to be routed through hybrid committees such as that on Defence Policy and Requirements set up in 1935 (Statement, 1936: 13). The Dominions Secretary was not even a member of this committee and was only appointed to the similar Defence Policy (Plans) committee in August 1939 when it had already ceased to meet (Roskill, 1974: 178; Chatfield, 1939b).

But then, the Dominions did not have, or want to have, any kind of locus standi on such important domestic issues. Ultimately, as all these practitioners of Westminster systems were well aware, such matters were exclusively the province of the ministers responsible to the relevant parliament. This, indeed, was the reason why Smuts could only attend, rather than become a member of the War Cabinet during the First World War, and why Menzies� request to join the War Cabinet during the Second was rejected.

This was also why the imperial dimension of CID functioned as it did. CID could not do otherwise and preserve the principle of accountability to respective parliaments which Borden, for instance, made so much of. More formal arrangements, even of the vague and limited nature envisaged by the IWC resolutions of 1917 and 1918, might have done violence to this constitutional requirement. This is quite apart from the political difficulties, in terms of allowing Britain and the Dominions a greater say in each others� affairs, that such arrangements would also risk. CID, in contrast, involved limited commitment both from Britain and the Dominions. There might have been complaints from the Dominions, especially in times of tension, over whether the amount of information being shared was sufficient. But generally a loose and limited relationship seems to have suited both Britain and the Dominions.

This may make it seem that the Committee of Imperial Defence was something of a misnomer, and that the imperial dimension of its work proceeded on a lowest common denominator basis. It was a planning organisation, but the imperial dimension did not really come into play for planning. Instead, the main factor binding Britain and the Dominions through CID was the need to share and receive information. Even here, however, there were limits to how far Britain and the Dominions were prepared to go, lest it should imply or involve commitments greater than their governments were prepared to enter into.

Imperial enthusiasts certainly felt this to be an unsatisfactory situation. A Royal Commonwealth Society [RCS] pamphlet argued in 1945 that the system of defence co-operation between the wars was not a success. Not the least problem, argued the great imperial federalist Lionel Curtis, was that too much of the burden remained upon the United Kingdom. However, to have shared the burden more widely would have required not only political will, which to Curtis seemed particularly absent in Canada (Curtis, 1945). It also would have required something more akin to a federal structure. Despite the prevalence of federalism within the Dominions, however, this was not something the RCS felt was likely to work on an empire-wide basis. Edwardian Liberal Unionists such as Lord Selborne may have had hopes in this direction (Boyce, 1987: xiii). The RCS, however, concluded, �The essential reason for the rejection of federal solutions is that they would limit too greatly and severely the spirit of liberty which is so dear to peoples of British stock� (Royal Commonwealth Society, 1945: 16).

Such quasi-mystical explanations may have their attractions for some people. And the defence of particular interests clearly was a key factor, both for Britain and the Dominions. So too was the operation of Westminster systems. Clearly Westminster systems can accommodate federalism, but a federalism within which defence is a reserved power. This was never likely to be an option in an world-wide system of imperial defence where the strategic interests of the contracting powers were so varied and overlapped (except in the highly significant case of Britain) so little. This is quite without considering the difficulties posed by the asymmetrical nature of the defence commitments of the governments involved. A federal system, for instance, could only weaken British control over its own defences, despite its preponderant contribution to the imperial defence effort. It could also threaten the Dominions not only with British interference but with commitments they neither desired nor wished to afford. In the circumstances, then, there is a much more obvious reason than that given by the RCS as to why federalism was not pursued within the empire; it was not conspicuously in anyone�s interest to do so.

The commitments involved in the operation of CID might seem minimalistic in comparison. They were indeed too minimalistic to justify the dignity of confederalism. The CID was essentially a British body with limited Dominion involvement. As an imperial body it was a deeply asymmetrical one. Confederalism, even of the limited kind envisaged by the IWC, was not appropriate. Instead, as has been shown, the associational modus operandi of CID broadly suited all the parties involved. It provided a channel for the information flows which both Britain and the Dominions required, without posing broader political problems. And yet this loose structure also succeeded in keeping the empire together whereas more formal ties probably would have had an opposite effect, at least as far as South Africa was concerned. Instead, the Dominions made a valuable contribution to the effort in two world wars. Therefore, not only was Attlee right when making a virtue of imperial flexibility. Measured by the outcomes in two global conflicts, notwithstanding the RCS�s opinion, CID succeeded.

Footnotes:
1 This paper derives from research for a project on the British Cabinet committee system funded under the Economic and Social Research Council's Whitehall Programme. I am grateful to the ESRC, award no. L124251002. *
2 It could also incorporate expertise from outside government (in the 1930s some technical committees consisted entirely of outside experts) or even from the Opposition benches, as both Balfour and Haldane were to prove. *
3 Australian elections in 1917 prevented any Australian representation. *
4 The devolved territory of Northern Ireland was also not represented in the CID system, defence being one of the powers reserved to Westminster. *
5 These figures are derived from tables given in PRO: CAB 21/2455 and CAB 21/2456. *
6 This is apparent from the CID committee book series which runs from 1927 to August 1939, PRO: CAB 59/1-6. *

 

References:
Attlee, C R (1942), letter to Churchill, 22 July 1942, Public Record Office, Kew [henceforward PRO]: CAB 104/180.
Boyce, D G (1987), The Crisis of British Unionism: The Domestic Political Papers of the Second Earl of Selborne, 1885-1922, (London: The Historians� Press).
Bridges, Sir Edward (1944), letter to Churchill, 17 May 1944, PRO: CAB 104/181.
Campbell, Sir Gerald (1939), letter to Sir Edward Harding, 20 January, PRO: CAB 21/2467.
Catterall, P P and C Brady (forthcoming), �Cabinet Committees in British Governance�, Public Policy and Administration.
Chatfield, Lord (1939a), letter to Inskip, 4 August, PRO: CAB 21/2464.
Chatfield, Lord (1939b), letter to Inskip, 23 August, PRO: CAB 21/2464.
Churchill, W S (1942a), note on bottom of India Office letter to Churchill, 6 June, PRO: PREM 4/43A/14.
Churchill, W S (1942b), note on bottom of Bridges letter to Churchill, 16 July, PRO: CAB 104/180.
Churchill, W S (1943), minute, 12 February, PRO: CAB 104/181.
Churchill, W S (1944), letter to Dominions Secretary, 3 February, PRO: PREM 4/43A/14.
Cranborne, Lord (1941), letter to Churchill, 18 August, PRO: PREM 4/50/4.
Crozier, A (1997), The Causes of the Second World War, (Oxford, Blackwell).
Curtis, L (1945), Comments, National Library of Wales, Aberystwyth, Dr Thomas Jones CH Papers, Class W, vol. 3, doc. 118.
Dominions Office (1942), 27 January, PRO: PREM 4/43A/14.
Elliott, W Y (1930), �A Written Constitution for the British Commonwealth�, Political Quarterly, 1, 1930, pp. 386-409.
Esher, Lord (1904), Report of the War Office (Reconstitution) Committee, Cd 1943.
Hankey, Sir Maurice (1922), �The Work of the Cabinet Secretariat�, 23 October, PRO: CAB 21/276.
Hankey, Sir Maurice (1927), letter to L Amery, 18 July, PRO: CAB 21/470.
Hankey, Sir Maurice (1929), memorandum, 22 May, PRO: CAB 21/470.
Hankey, Sir Maurice (1930), �The Association of the Dominions with the Crown�, 21 November, PRO: CAB 21/470.
Hankey, Sir Maurice (1936), �The Higher Aspects of Cabinet Control of the Armed Forces of the Crown�, PRO: CAB 21/736.
Hankey, Sir Maurice (1937), letter to C T te Water, 29 May, PRO: CAB 21/472.
Ismay, H L (1938), letter to N Chamberlain, 17 September, PRO: CAB 104/123.
Ismay, H L (1939) �The Machinery of the Committee of Imperial Defence�, RUSI Journal, May.
Johnson, F A (1960) Defence by Committee: The British Committee of Imperial Defence 1885-1959, (London: Oxford University Press).
Keith, A B (1942), letter to The Times, 3 February.
Laski, H (1939), �The Structure of the War Cabinet�, New Statesman and Nation, 11 November, pp. 669-71.
Macintyre, S (1986), The Oxford History of Australia, Vol. 4: 1901-42, (Melbourne: Oxford University Press).
Mackay, R F (1985) Balfour: Intellectual Statesman, (Oxford: Oxford University Press).
Minutes of meeting with Canadian ministers (1943), 31 August, PRO: PREM 3/366/11.
Otte, Thomas (1996), �A Question of Leadership: the Unionist Cabinet and Foreign Policy 1895-1900�, unpublished seminar paper presented at the Institute of Historical Research, London, 20 March.
Roskill, Stephen (1972), Hankey: Man of Secrets, Vol. II, (London: Collins).
Roskill, Stephen (1974), Hankey: Man of Secrets, Vol. III, (London: Collins).
Rowan, C L (1945), letter to Brook, 24 March, PRO: CAB 104/181.
Royal Commonwealth Society (1945), The Political Future of the British Commonwealth and Empire, 25 May.
Statement Relating to Defence (1936), Cmd 5107.
War Cabinet (1917), Report for the Year 1917, Cd 9005.
War Cabinet (1918), Report for the Year 1918, Cmd 325.
Whiskard, Sir Geoffrey (1941), letter to Cranborne, 18 June, PRO: PREM 4/50/4.
Yates, G (1939), letter to Tucker, 3 September, PRO: PREM 1/384.