Words: 9891
Jari Eloranta, Ph. D Assistant Professor of Comparative Economic and Business History, Appalachian State University, Department of History, Whitener Hall, Boone, NC 28608, USA Phone: +1-828-262 6006, email: elorantaj@appstate. edu Paper to be presented at the Sixth European Historical Economics Society Conference, 9-10 September 2005, Historical Center of the former Imperial Ottoman Bank, Istanbul. WHY DID THE LEAGUE OF NATIONS FAIL? INTRODUCTION
The economic and political instability of the interwar period and the rise of authoritarian regimes are often seen as extensions of World War I and the Great Depression. The League of Nations, in turn, is usually seen as an organization that failed to act adequately during the various political crises of the period, beginning with the Japanese aggression in Manchuria. But, I would argue that its failure has to be seen in the larger context of the failed disarmament processes of the interwar period. Why did the League of Nations ultimately fail to achieve widespread disarmament, its most fundamental goal?
Maurice Vaisse (1993) has summarized the explanations in the following manner: 1) It failed because it was an imperfect instrument for achieving disarmament; 2) It failed because the League was not universal; 3) It failed because of the confrontation between Great Britain and France; 4) It failed because there were domestic forces inside the countries hostile to disarmament; 5) It failed because the Disarmament Conference was convened too late, under hostile conditions; 6) It failed because of the confrontation between France and Germany at the Disarmament Conference; 7) It failed because of the overly ambitious aims and the practical problems involved in the reduction of armaments. 1 And, as Frederick Northedge has argued, the League failed because it was seen as the defender of the status quo, the infamous Versailles settlement. 2 As argued here, all of these explanations have merit, yet the list 1 2 Vaisse 1993. Northedge 1986, 288-289. 2 is hardly exhaustive. First, contrary to Vaisse, I would maintain that the disarmament that took place contemporaneously in the 1920s and the early years of the Great Depression did not offer a real window of opportunity for disarmament. Second, the role of the “weak” states was not as constructive as is often perceived, since they could not offer a unified front on most issues. Nor were they all pacifistic in the vein of the Scandinavian countries. Third, as argued in Eloranta (2002b), the domestic opposition among economic interest groups to, for example, arms trade regulation was quite formidable. Finally, the rigid negotiation stances of the key states in the disarmament process prevented a more favorable outcome, since far-reaching compromises were required from all participants. Thus, the states tended to pursue their own interests, which were not the same for each state nor were the means that they were ready to use to achieve their aims.
The way that these interests emerged in the foreign policy of a particular state was a combination of external (systemic, alliance-specific, dyadic) and internal (economic, political, actor-specific) factors. 4 I would argue that the failure of the League of Nations had two important dimensions: 1) The failure to provide adequate security guarantees for its members (like an alliance), thus encouraging more aggressive policies especially by the authoritarian states and leading to an arms race; 2) The failure of this organization to achieve the disarmament goals it set out in the 1920s and 1930s, such as imposition of military spending constraints.
These dimensions, including the aggregate explanations of the weaknesses of the League of Nations, have not been explored adequately by the extensive literature on the interwar economic and political turmoil. I would argue that analysis of these failures by the League of Nations can increase our understanding of the military rivalries, regime changes, and, ultimately, the outbreak of World War II. First, here I will analyze how and why the League of Nations failed to provide credible security guarantees during the interwar period, and what this failure meant for the military spending decision-making of the member nations. The foreign policy Cf. Vaisse 1993, 185-186. On this type of argument, see Rosecrance and Stein 1993; Rosecrance and Steiner 1993, 124???125.
More specific tests of this argument, in relation with military spending demand and the relevant variables, are presented in the subsequent sections. 4 3 3 environment under the superficially strong League of Nations in the 1920s did not provide encouragement for meaningful spending cuts. Moreover, what did the League of Nations Covenant actually propose in terms of security and how did the different players adapt to this framework? I will also explore the multitude of efforts to achieve credible disarmament measures5, from the early 1920s to the early 1930s. In fact, did the League of Nations fail to provide the right institutional setting for the disarmament bargaining or was it doomed to fail, due to inadequacies related to its structure and the players involved? The evidence uncovered in this paper suggests that it was doomed to fail, given the inability of the League to make credible security guarantees and the widely differing goals of the members. Second, I will explore whether the League of Nations actually could be modeled as a credible (or indeed failed) military alliance, i. e. whether the military spending of the sample members exhibited pure public good characteristics. Beginning with Mancur Olson and Richard Zeckhauser’s path breaking work on NATO7, there have been many testable hypotheses relating to the idea of collective security provision in an alliance and the implications of this provision on military spending.
As the logic suggested by Olson and Zeckhauser implies, pure public good alliances are characterized by free riding by the smaller (or poorer) states. For example, more recent studies have found the pure public good alliance concept to describe the NATO until 1966, when a change in the strategic doctrine forced the members to rely more on their own military provision. 8 This paper thus explores whether free riding occurred, with both time series and cross section samples based on Eloranta (2002b), in order to assess the public good qualities of the League of Nations as an “alliance”. It seems that League of Nations did not function as a credible alliance.
Member states by and large pursued their own military readiness solutions. Ultimately, thus, the question is: If the League failed to provide pure public good in military deterrence, is this the reason why this organization failed? The evidence in this paper suggests that See e. g. Barros 1993. For a more complete archival analysis of the negotiations and the players involved, as well as discussion of the data and sources, see Eloranta 2002. 7 Olson and Zeckhauser 1966. 6 5 4 this was indeed the case. It failed in its most fundamental goal, namely to be able to guarantee that peace would last. COLLECTIVE SECURITY ASPIRATIONS AND THE LEAGUE OF NATIONS
The aim of this section is to undertake a review of the functions of the League of Nations, especially in terms of aspirations towards disarmament and a general strive towards improved collective security. Also, I will briefly review some of the key measures aimed at promoting peace outside the League of Nations. First I will analyze the structure and functions of the League of Nations in the sphere of collective security, by necessity often at an elusive macro-level. Finally, I will conclude this section by reflecting on the impacts of American isolationism for Europe, and the broad policy stances of the United Kingdom and France in comparison.
The termination of the First World War on November 11, 1918 brought with it a quick withdrawal of American forces from Europe and a significant initial change in the political attitudes of the Great Powers. With the devastation brought on by the war, the advancement of peace became a popular theme in international politics in the 1920s. 9 The efforts to achieve peace rested on Wilson’s Fourteen Points, developed on the basis of many similar ideological schemes that surfaced during the Great War, and the Treaty of Versailles, which included the foundations of the League of Nations, to bring the United States back into the policy of internationalism. 10 These aspirations were dealt a severe blow right from the beginning. For many reasons, the attempt to bring the United States back into the international politics failed. The U. S.
Senate refused to ratify the treaty first in November 1919, and again in March 1920 as Woodrow Wilson stubbornly refused to compromise. 11 Thus, Wilson and the League of Nations were defeated politically in the United States. Warren G. Harding’s victory in the American 8 9 Sandler and Hartley 1999. Preston and Wise 1970, 78-79; Pogue 1963. 10 Eloranta 2002b; Soule 1989, 81-82; Northedge 1986, 25-30. 11 Eloranta 2002b; Bemis 1959, 437-439. 5 presidential elections in 1920 led the United States into isolationism again, which became the leading guideline in American foreign policy for two decades. 12 The League of Nations came into existence on January 10, 1920.
All in all eighteen states became members of the League at first by formally approving the peace treaty. By late 1920, the number of members had already grown to over forty. In 1938, by comparison, the member count was fifty-five. The member states of the League of Nations in 1920 comprised 74 per cent of the world’s population and, respectively, 63 per cent of the world states’ area. 13 If one were to ask what the League of Nations was, Zara Steiner’s depiction might be considered fairly accurate: “It was an institutionalized form of collective action by the sovereign states to maintain the peace. “14 The League also became the supreme defender of the Versailles settlement in the postwar world.
And, it represented no real effort to create a supranational authority, because its decision-making structure granted the member states the final say in all matters. 15 The League of Nations’ structure consisted of three essential bodies: 1) Assembly; 2) Council; 3) Secretariat. The first two were the ones that had the power to act, whereas the Secretariat formed the functional bureaucracy of the League. The Assembly was in essence the League’s legislative arm, whereas the Council functioned as its cabinet. The Assembly consisted of not more than three representatives of each member state, and the Council had eight members (the ninth seat was reserved for the United States). 16
The Assembly was meant to assuage the fears of the smaller states, since the number of representatives was set to be equal among members. The Assembly also exercised considerable control, when it wished to do so, over the Council ??? it, for example, controlled the appointment of four out of eight Council members as well as determined its size and basic character. The Council, in turn, was intended to give the smaller states even further say in the matters of the League, because four of the eight represented the 12 13 Bemis 1959, 440-441. The League of nations starts; an outline by its organisers 1920, 1-5; Northedge 1986. 14 Steiner 1993, 38. 15 Northedge 1986, e. g. 288-289; Steiner 1993, 41-43. 16 See more Northedge 1986. 6 maller states (Belgium, Brazil, Greece, and Spain), yet they were only meant to be temporary seats. Permanent seats were occupied by the former Allied Great Powers: the United Kingdom, France, Italy, and Japan. By design, the Council was expected to be (which turned out to be true) a passive organization, although most policy decisions were up to the Council to initiate and act upon. Finally, the organization was run by a Secretary-General, with the Secretariat at his disposal. 17 The cornerstone of the League of Nations was its Covenant, consisting of 26 articles, which remained mostly intact in the form adopted in 1919 throughout the interwar period. 18 Articles 1-7 set up the central machinery of the League.
The basic premises of the disarmament were outlined in Articles 8-9, and Articles 10-17 elaborated on the socalled League system for the prevention of war, cited by the organizers as “the one great object of the whole organization”. The rest of the Articles, “following the piece de resistance”, dealt with a miscellaneous group of “important matters”. 19 The Covenant’s Article 8 consisted of the key elements of the future disarmament. It maintained that the League Council was to assess each state’s security needs and to formulate a plan for the disarmament of its members, based on this assessment. Equally, with this Article and others the League was entrusted with the task of supervising the collection of information on the development of armaments among its members. 0 It maintained that “the Members of the League recognise that the maintenance of peace requires the reduction of national armaments to the lowest point consistent with national safety and the enforcement by common action of international obligations. ” Additionally, it put the Council in charge of achieving this goal: “The Council, taking account of the geographical situation and circumstances of each State, shall formulate plans for such reduction for the consideration and action of the several Governments. “21 Moreover, Article 9 provided for the establishment of a permanent commission, The Permanent Advisory Commission for Military, Naval, and Air Questions, to study what criteria to The League of nations starts; an outline by its organisers 1920, 5-18; Northedge 1986, 49-55.
Compare: The League of nations starts; an outline by its organisers 1920, Appendix I (containing the 1919 Covenant) and Northedge 1986, Appendix A (containing the 1938 amended Covenant). Many of the amendments dealt with changes in the administrative structure of the League. 19 See the The League of nations starts; an outline by its organisers 1920, 26; Appendix I 20 Barros 1993, 606-607; Peterson 1993. 21 The The League of nations starts; an outline by its organisers 1920, Appendix I, Article 8. 18 17 7 set for the member states and to request relevant information pertaining to these matters. This commission was also open to representatives of all states, including non-League countries when the matters were of concern to them. 22
The Articles relating to the resolution of conflicts, through voluntary or forced arbitration, were both impressive in their detail yet vague in their enforcement. It provided for conflict resolution both among the League members, as well as between a League member and non-member. The general principle was, outlined in Article 11, that “any war or threat of war, whether immediately affecting any of the Members of the League or not, is hereby declared a matter of concern to the whole League, and the League shall take any action that may be deemed wise and effectual to safeguard the peace of nations. “23 The principles of arbitration were found in Articles 12, 13, and 15.
Article 16, the most important part of the Covenant in terms of collective security guarantees, outlined the courses of action available to the League and its members in case arbitration was not successful or it was rejected altogether:24 “Should any Member of the League resort to war in disregard of its covenants under Articles 12, 13, of 15, it shall ipso facto be deemed to have committed an act of war against all other Members of the League, which hereby undertake immediately to subject it to the severance of all trade and financial relations, the prohibition of all intercourse between their nationals and the nationals of the covenant-breaking State, and the prevention of all financial, commercial, and personal intercourse between the nationals of the covenant-breaking State and the nationals of any other State, whether a Member of the League or not. It shall be the duty of the Council in such case to recommend to the several Governments concerned what effective military, naval, or air force the Members of the League shall severally contribute to the armed forces to be used to protect the covenants of the League. As was the case with the advancement of disarmament, the responsibility for determining the necessity of sanctions, or indeed a harsher punishment, was placed on the hands of the Council. Nonetheless, the Council’s task was merely to recommend what contributions member states should make to the armed forces to protect the Ibid. , 137-138. See also the later Temporary Mixed Commission for the Reduction of Armaments, e. g. its report in League of Nations, Assembly Documents, A. 35. 1923. IX (Part 1): August 30. 23 Ibid. , Appendix I, Article 11. 24 Ibid. , Appendix I, Article 16. 22 8 covenants of the League. 25 The same standards were applied to both intra-League and extra-League conflicts.
Article 17 stated that “in the event of a dispute between a Member of the League and a State which is not a Member of the League…If a State so invited shall refuse to accept the obligations of membership in the League for the purposes of such dispute, and shall resort to war against a Member of the League, the provisions of Article 16 shall be applicable as against the State taking such action. “26 Finally, Article 23 trusted the League with the supervision of the trade in arms and ammunition as well. 27 Thus, the war-prevention strategy of the Covenant consisted of four parts: 1) Reduction of the means to fight wars with; 2) Enforcement the principle of sending unsettled disputes to third-party settlement and taking measures against states which refused the arbitration or ignored its outcome; 3) Exchange of guarantees on the status quo created by the Treaty of Versailles; 4) Resolution of international conflicts before they become dangerous for world peace. 28
It became apparent early on in the 1920s that individual states, regardless of their commitment to the League framework, would not disarm unless they felt secure. This for most states meant some type of a framework of collective security guarantees. Within the League of Nations, the member states attempted to achieve disarmament measures in earnest at least until the mid-1930s. The first plans right after the war were merely aimed at containing the military spending levels or preferably reducing them. 29 All of these efforts failed due to one reason or another (for example, due to different conceptions of disarmament among the Great Powers). Mostly these efforts took place between diplomats within the various committees in the League of Nations’ political machinery.
One of the many difficulties with achieving concrete results was the heterogeneous nature of the participants and their different expectations. For example, the Geneva Protocol of 1924 ??? which advocated principles such as denouncing of war, agreement on sanctions against aggressor(s), and the convening of a disarmament Northedge 1986, 56-57. The League of nations starts; an outline by its organisers 1920, Appendix I, Article 17. 27 Ibid. , Appendix I, Article 23. 28 Northedge 1986, 54. See also Steiner 1993. 29 See e. g. League of Nations, A. 13. 1921. IX: August 22. Reduction of National Expenditure on Armaments. Including replies from 26 governments; League of Nations, C. 90. M. 40. 1921. IX: June 7.
Reduction of National Expenditure of Armaments. Including replies from 16 governments. 26 25 9 conference ??? failed ultimately due to British rejection of the Protocol. In 1925, nonetheless, a Preparatory Commission for the Disarmament Conference was established, to prepare for a disarmament conference. Its work took five years and culminated in the Disarmament Conference of 1932-1934. 30 The Preparatory Commission was not particularly successful in its endeavors, since there were, for example, major disagreements between the British and the French over the scale of disarmament and security guarantees, and between the British and the Americans over naval disarmament.
There were three important sources of disagreement still left concerning the commission’s recommendations before the actual conference took place: 1) It made no provision for the inclusion of trained reserves; 2) It assigned no limitations for the materials of armies and navies; 3) There were no restrictions placed on the cost of material for the air forces. 31 The Conference for the Reduction and Limitation of Armaments was convened on February 2, 1932, when the League was already experiencing the difficulties of furthering world peace due to Japanese aggression against China in Manchuria. Before the opening, a draft resolution of the Danish, Norwegian, Dutch, Swedish, and Swiss delegations was submitted requesting the Council to urge governments to abstain from increases in their level of armaments.
These countries had made a similar suggestion already on September 11, 1931, leading to the acceptance of this principal by the participants of the upcoming conference. This was a curious solution, one of the few temporarily successful disarmament acts, known as the armaments truces, in fact very similar to the early armaments containment efforts in the 1920s. Thus, an armaments truce was maintained during the Disarmament Conference, which of course was the peak of the Great Depression for many countries, making the acceptance of this measure less challenging. 32 Although the conference met on and off for nearly two and a half years, it failed to produce tangible results. The differences of opinion between the Barros 1993, 615-619; Northedge 1986, 56; 118-119.
Northedge 1986, 120-121. See also Scott 1973, e. g. 189. A good overall account of the disarmament “process” can be found in Vaisse 1993. 32 See e. g. League of Nations, A. 51. 1931. IX: September 11. Reduction of Armaments; League of Nations, C. 774. M. 369. 1931. IX: October 29. Voluntary Armaments Truce (Czechoslovakia’s positive reply); League of Nations, C. 627. M. 309. 1932. IX: September 5. Voluntary Armaments Truce. Replies from 6 governments (concerning the extension of the armaments truce). See also Jones 1939, 244-246. 31 30 10 participants were simply too great, and the emergence of Hitler’s rule in Germany at this time hardly helped matters.
The failure of the League in Manchuria dealt the League an “almost fatal blow”, surely also contributing to the failure of the Disarmament Conference. On the spring of 1934, the Danish, Norwegian, Spanish, Swedish, and Swiss delegations made a final plea to overcome the impasse of the conference, but to no avail. Hopes of collective disarmament died in a definitive fashion with the departure of Germany from the League. Fifteen years of effort had been wasted. When the Disarmament Conference met for the last time on June 11, 1934, the disarmament “process” was effectively dead on its tracks. 33 One of the few successful disarmament undertakings by the League of Nations was the collection of data on military spending and arms trade.
The first report, titled Budget Expenditure on National Defence: 1913 and 1920???1922, was published in 1922, and it included military spending data for twenty-one countries in the post-war period. The introduction to the report made specific stipulations on the limitations of the enquiry, yet the data for each country also included real value comparisons with the prewar period. The aim was “to furnish material indicating the development and tendency in each individual country”. 34 The next report was issued already in 1923, titled Statistical Enquiry into National Peace-time Armaments. This time the report included two volumes, nearly two hundred pages in total. Also, the report gave detailed data on the military spending, including the colonies, and the armed forces of seventeen nations in all. 5 One of the most detailed statistical sources on peacetime armies and military spending of the world was created on the same principles in 1924, as the Armaments Year-Book came into existence. 36 This yearbook continued to provide fairly reliable data on the military establishments of most countries, excluding Germany in the late 1930s, as I have already discussed in Eloranta (2002b). Ibid. , 249-251; Northedge 1986, 122-134; Vaisse 1993. League of Nations, A. 38(a). 1922. IX. Budget Expenditure on National Defence: 1913 and 1920??? 1922, 1???5. See also League of Nations, Armaments Year-Books (1924???1940). 1st Edition. A. 37. 1924. X; (1940 Edition) C. 228. M. 155. 1939. IX. 35 League of Nations, A. 20. 1923.
IX, Statistical Enquiry into National Peace-time Armaments, Parts 1 and 2. 36 See Eloranta 2002b, Appendices, Appendix 2. 34 33 11 In the period following the First World War, there were a multitude of efforts aimed at controlling and limiting the trade of armaments with the League of Nations. Among others, the Versailles Peace Treaty established controls for the arms trade of the losers of the war, and there were negotiations on embargoes and production limitations both at the national and the supranational level, mainly under the auspices of the League of Nations. Similarities with the overall disarmament process are abundant. For example, the St.
Germain Convention for the Control of the Trade in Arms and Ammunition of 1919, which attempted to establish a government licensing system for certain weapons, was to be supervised by a specific League of Nations’ Central International Office. The downfall of this agreement was mainly the opposition of the United States to the close connection with the League of Nations. A broader agreement following the Conference for the Supervision of the International Arms Trade in 1925 ran into similar problems when it came to the ratification phase of the convention. The participating states attached conditions to the ratification process, which effectively destroyed its chances of success. The United States’ Senate did not ratify the convention until 1934.
The main results that emerged from the 1920s armaments trade limitation negotiations were the establishment of a licensing system among the nations, which also served as a basis for the gathering of statistical information on this type of trade in the League publications. One of the basic aims in the control efforts of the 1920s and 1930s was to ban private manufacture and sale of armaments, although these initiatives never really produced concrete results, with especially the ratification phase usually providing the final blow to any regulative endeavor. 37 What was the significance of extra-League peace initiatives, with usually a larger participation base and more precise focus in the negotiations? The first significant peace conference, used here as an example, took place in the aftermath of the First World War and the failure of the American involvement in the League of Nations.
The Washington Conference on Naval Limitation became a reality in November 1921, with participation by the United States, Great Britain, Japan, France, Italy, Belgium, Netherlands, Portugal, and China. The question of Far East, mainly China, was one of the important 37 Krause and MacDonald 1993, 713-717. See also Krause 1992. 12 issues at the conference. Since the Americans were not ready to fight for China, the treaties provided an excuse not to carry the burden of the protection of the Far East. 38 The most important of the treaties accomplished in the Washington Conference was the Washington Treaty for Limitation of Naval Armaments in 1922.
In this treaty the United States, Great Britain, and Japan (in addition to France and Italy) agreed upon maintenance of a battleship and carrier ratio of 5-5-3 for the “big” naval powers (and 1. 7 respectively for the others). 39 With the Four Power Treaty (1921), the United States, Japan, Great Britain, and France agreed upon displaying mutual respect for their interests and possessions in the Pacific region. The third treaty, the Nine Power Treaty (1922), was designed especially to solve the problem of China’s “defense”, which in practice meant the continuance of Western domination. In this treaty the nine naval powers agreed, at least in principle, to respect China’s sovereignty, independence, and other matters. 0 Even though these reforms surely had honorable aims and were recognized to be outstanding achievements in their time, the outcomes proved to be disappointing in many ways. The limitations were not always obeyed. 41 The treaties accomplished in the Washington Conference also lacked the machinery required to enforce the established agreements. For example, the Nine Power Treaty on China was mainly rhetorical by nature and did not offer anything concrete for its enforcement. 42 Yet, when did the League’s impotence in the task of maintaining the world peace become apparent to all of its members? In fact, it was the weaknesses that were contained in the League framework and the foreign policy stances of the members that made it impossible for the system to work.
The real test of the covenants first came with the surprising Japanese aggression in Manchuria. This turned out to be quite a shock for the League members, since Japan, a permanent member of the Council, had been conciliatory in its foreign policy in the 1920s, even during the naval disarmament conference of 1930. The long slide into war in Manchuria began on September 18, 1931, when local Japanese army attacked the city of Mukden without the knowledge 38 39 Bemis 1959, 457-460; Eloranta 2002b. Eloranta 2002b; Bemis 1959, Peterson 1993. 40 Eloranta 2002b; Northedge 1986. 41 Hicks 1960, 39, 41-43, 49; Eloranta 2002b. 13 and the wishes of the government in Tokyo.
The government was forced to follow the military’s lead in the matter, and the incident developed into an international conflict as the Japanese made considerable headway against the inferior Chinese forces. This prompted extensive debate in the Council, yet it was not willing to put heavy pressure against Japan. Further Japanese military action in Shanghai on January 28, 1932, finally triggered a more unitary collective response, which, despite being quite cautious, got Japanese troops out of Shanghai. When a special report condemning Japan was approved on February 24, 1933, the Japanese delegation walked out of the Assembly. Japan announced her formal withdrawal from the League on March 27, 1933. 43 The “Manchurian Incident”, as it was called, was just the first of many deadly blows to come at the League of Nations.
The Soviet Union’s joining of the League in 1934 at first provided a signal of hope for peace. Hitler’s ascendancy to power in 1933 and his revisionist ideas soon came to fore in European politics. Germany’s withdrawal from the League (admitted to the League in 1926) and its fevered rearmament from 1935 onwards certainly cast doubts on the League’s and Europe’s future. Equally, the process of “peaceful” conquests started by the remilitarization of Rhineland in March 1936, leading up to the Second World, were certainly among the death blows to the League’s credibility. 44 Yet, inability of the League to halt Italian (another member of the League’s Council) aggression in Abyssinia in 1935-1936 turned out to be its most decisive fiasco.
Mussolini, in essence, was able to achieve his illicit conquest despite the protestations of the other European and world powers. Especially the British, who initially were the prominent force behind them, were against the continuation of the sanctions put in place under Article 16 initially, and thus even the sanctions were removed in July 1936. This merely acknowledged the prevailing situation: the Great Powers were not ready to initiate aggression against Italy due to this conflict, and that Mussolini’s victory in Abyssinia had already been sealed months before. To many revisionists, especially 42 43 Hicks 1960, 46-47. See e. g. Scott 1973, 208-229; Northedge 1986. 4 See, for example, Murray 1984; Kennedy 1989; Northedge 1986. 14 Hitler, this meant that the League was truly unable to stand in the way of the redrawing of the map of Europe and the destruction of the status quo created at Versailles. 45 Moreover, the American isolationism, however inadequate as the term may be, left the European and even the “world” power relations in the hands of Great Britain and France. They were reluctant leaders in their own right, with their own interests displayed in their actions for example in the League of Nations functions. Germany and Russia had been defeated in the First World War, thus leaving room for these traditional Great Powers to re-emerge in European politics.
There were obvious disagreements in the goals valued by the British and the French. The British, like the Americans, were less and less interested in the goal that France valued the most: keeping Germany in check. Additionally, Great Britain was more pre-occupied by extra-European problems, namely keeping the vast Empire from disintegrating. At the beginning of the 1930s France seemed to be the leading nation in the European scene. Its economic performance in the 1930s, however, proved to be poor in comparison with the other European Great Powers. 46 Thus, the European stage created a sort of a “power vacuum” during the 1930s, which seemed to invite hegemonic competition for leadership. BUT WAS IT EFFECTIVE?
MILITARY SPENDING AND NAVAL BUILDUP IN THE INTERWAR PERIOD After the First World War, especially in the 1920s, although the defense shares (=ratio of military spending to central or federal government expenditures, %) of large democracies dropped noticeably, their respective military burdens (=ratio of military spending to GDP, %) stayed either at similar levels as before or actually increased. As has been noted in Eloranta (2002), the mean military burden of 17 countries47 did not change much from the period 1870???1913 to the interwar period (1920-1938); in fact, only a slight increase occurred. However, the mean defense share of the pre-First World See especially Northedge 1986, Chapter 10. Kennedy 1989, 357-375; Pearton 1982, 177, 197-198. 7 Includes: Austria, Belgium, Denmark, Finland, France, Germany, Italy, Japan, the Netherlands, Norway, Portugal, Russia/USSR, Spain, Sweden, Switzerland, the UK, and the USA. 46 45 15 War period was nearly double that of the interwar period, with the latter amounting to eighteen per cent. Thus, the mean military burdens proved quite resistant to change (or path dependent), whereas the military expenditures’ (abbreviated: ME) budget shares shrunk noticeably as new spending programs were introduced in the spheres of social issues and education. For example France’s military burden average increased respectively in the interwar period compared to the earlier period, and its mean defense share declined only slightly.
Other countries that behaved similarly included Belgium, Portugal (with also its mean defense share increasing after the First World War), Spain, and the UK. Figure 1. Military Burdens (=Ratio of Military Expenditures to GDP, %) of Germany, Italy, Japan, and Russia/USSR, 1920-1938 % 25 20 15 10 5 0 1920 1922 1924 1926 1928 1930 Year 1932 1934 1936 1938 GER, MILBUR JAP, MILBUR ITA, MILBUR RUS, MILBUR Sources: see Eloranta (2002b) for details. The military burdens of the authoritarian and totalitarian challengers of France and the United Kingdom grew strongly from 1933-1934 onwards, and the overall levels of these countries during the authoritarian rule were in general higher than those of most democratic states.
One should also take note that the more authoritarian the nation was 16 the higher its military burden seemed to be. For example, Mussolini’s Italy, with Mussolini unable to subjugate and consolidate all societal groups under his direct rule, seems to have been unable to match the militaristic drive of Japan (under a military regime) and Germany (under Hitler’s centralized regime), perhaps also the Soviet Union, in the late 1930s. For the European democracies, the mid-1930s in general marked the beginning of rather slow rearmament, although their authoritarian challengers had begun earlier. Hitler’s Germany increased its military burden from 1. 6 per cent in 1933 to 18. in 1938, a rearmament program promising both “guns and butter”. Mussolini’s efforts in Italy were less successful, producing a military burden of four to five per cent in the 1930s. The Japanese rearmament drive was perhaps the most extensive, relative of its economic base, amassing a military burden as high as 22. 7 per cent in 1938. 48 Thus, the League of Nations certainly was not able to contain the arms race of the 1930s, and had only a limited impact even on the 1920s’ development. What about the navies? After all, this was one of the areas in which the League was its most active. These questions require some estimations of the available military stock by the states to be compared.
One measure, advocated by George Modelski and William R. Thompson (1988)49, is the number of battleships, perceived by these authors to reflect the ability of a state to assume a leadership position within a system. As seen above in Figure 2, the number of battleships reflects rather well the earlier discussion of naval limitation agreements by the Great Powers. By 1931, the United Kingdom and the United States seem to have achieved a balance with the other states also maintaining steady numbers of battleships. Was the rearmament of the 1930s an illusion, especially in the sphere of naval armaments? Eloranta 2003 and the sources in it. See also Eloranta 2002b.
Modelski and Thompson also use somewhat scattered estimates of the Great Power naval spending to calculate world leadership shares as well as other variables. Their study is, in particular, to be commended due to its detailed explanations on the sources used and the weaknesses of the estimates. See Modelski and Thompson 1988, e. g. 38-48. Here I have chosen to define a battleship as a military capital ship other than an aircraft carrier with a tonnage of at least 15 000 tons. On concepts and theoretical challenges, see especially Modelski and Thompson 1996. 49 48 17 Figure 2. Number of Battleships by France, Japan, the United Kingdom, and the United States, 1920-1938 NUMBER 45 40 35 30 25 20 15 10 5 0 1920 1922 1924 1926 1928 1930 Year 1932 1934 1936 1938
FRA, BATTLESHIPS UK, BATTLESHIPS JAP, BATTLESHIPS USA, BATTLESHIPS Sources: see Eloranta (2002b) for details. It is argued here that this picture may be quite misleading as a representation of the naval competition, or the lack of it, in the 1930s. I have constructed figures on the depreciated total tonnages of the seventeen states, using the guidelines of the League of Nations on the depreciation lengths of different kinds of ships. As this procedure is extremely labor-intensive, the depreciated tonnages were constructed only for the years 1923, 1928, 1933, and 1938. The totals were then interpolated using the indices explained in Eloranta (2002).
These figures should provide a better estimation of the “true” naval stock of these nations, especially in terms of naval competition, because: 1) Battleships represent perhaps merely the offensive capabilities of states, or the ability to maintain “leadership”; and 2) Outdated materials were indeed deemed to be useless in battle, as displayed by the British estimations that during the First World War an older standard German battleship would last no more than five minutes against a modern, 18 British Dreadnought. 50 Let us first examine the comparative depreciated naval tonnages of some of the Great Powers in this period (see Figure 3 below). Figure 3.
Total Depreciated Tonnages of France, Germany, the United Kingdom, and the United States, 1920-1938 TONNAGES, IN BILLIONS 0. 0012 0. 0010 0. 0008 0. 0006 0. 0004 0. 0002 0. 0000 1920 1922 1928 1930 1932 1934 Year FRA, TOTAL DEPRECIATED TONNAGE GER, TOTAL DEPRECIATED TONNAGE UK, TOTAL DEPRECIATED TONNAGE USA, TOTAL DEPRECIATED TONNAGE 1924 1926 1936 1938 Sources: see Eloranta (2002b) for details. The view provided by the number of battleships seems to, in fact, have been quite accurate for the 1920s. The naval limitation agreements indeed decreased the usable tonnages of the United Kingdom and the United States, while the tonnage held by France in effect increased, producing an actual parity between the three.
This had not been the aim of the naval limitation accords discussed earlier. Nonetheless, quite divergent trends emerged after 1933. While the British rearmament programs, often in connection with depression-related employment efforts, produced a strong increase in the 1930s (with the U. S. effort somewhat more meager), the French actual naval stock declined, both due to lack of funding and the aging of the ships. The German fleet, practically nonexistent before the 1930s, was built up quite fast, at least to provide a 50 Modelski and Thompson 1988, 76. 19 significant threat to the French, yet the naval lead of the United Kingdom and the United States remained clear.
Figure 4 includes three possible indicators of naval threats among seventeen states. The first is a combined volume index constructed from the figures provided by Modelski and Thompson (1988) as well as other sources found in Eloranta (2002b), the second is the total nominal tonnage of the seventeen states, and the third is the total depreciated tonnage of the said states. This third index seems to indicate most clearly a decline in the 1920s, and an emerging growth trend in the 1930s. The combined volume index of “threat” ignores some of the disarmament tendencies of the 1920s, and the nominal tonnage indicates, for example, a more abrupt growth trend from the mid-1930s onwards.
I would argue that the total depreciated tonnage confers a more accurate picture of this phenomenon. All in all, disarmament did seem to occur in a limited sense in the 1920s, yet the 1930s devolved into an arms race once again. Figure 4. Index of Naval Threat (1928=100), 17-country Total Nominal Tonnage, and 17-country Total Depreciated Tonnage, 1920-1938 NAVAL THREAT INDEX TONNAGES, IN BILLIONS 300 250 200 150 0. 007 0. 006 0. 005 0. 004 0. 003 100 50 0 1920 0. 002 0. 001 0. 000 1938 1922 1924 1926 1928 1930 Year 1932 1934 1936 COMBINED INDEX (NAVAL SPENDING + BSHIPS) (1928=100) 17-COUNTRY, TOTAL TONNAGE 17-COUNTRY, TOTAL DEPRECIATED TONNAGE
Sources: see Eloranta (2002b) for details. 20 MILITARY SPENDING AS A PURE PUBLIC GOOD: THE LEAGUE OF NATIONS AS AN ALLIANCE? Here in this section I will assess the importance of pure public good characteristics for the military spending behavior of a selected sample, due to data concerns, of eleven European states: Belgium, Denmark, Finland, France, the Netherlands, Norway, Portugal, Spain, Sweden, Switzerland, and the United Kingdom. I will explore the idea that the League of Nations might have failed to produce a public good in the form of collective security, and what this failure entailed for the military spending behavior of these nations.
The military spending framework under the superficially strong League of Nations in the 1920s did not provide encouragement for meaningful spending cuts, which fits well within the proposed arms race model introduced in this section. Furthermore, the reasons behind the disarmament, the most important function of this organization, failure of the League are explored in this section, and especially the role of the “weak” states is re-evaluated. As the earlier discussion implied, the League of Nations can be argued to have formed an alliance of sorts, even in the military sense. An attempt was certainly made to provide a framework for collective security in the League Covenant, which consisted of various measures meant to force states into arbitration over disputes.
Furthermore, for example Article 16 provided measures, such as the assembly of a collective military force, for the enforcement of these principles. Bruce Russett has qualified the League of Nations as a “quasi-global collective security arrangement…which bind[s] all members to coalesce against any aggressor, even one of their own number”. 51 Despite the fact that the structure of the League and the different views of its members on these commitments rendered these ideals quite unreachable52, the League certainly on paper filled the requirements of a military alliance. And, as such, it can be investigated as providing a possible pure public good deterrence to its members. 53 The maintenance of collective security, i. e. t least some form of lasting peace, was indeed “the one great Russett 1971, 263. See especially Stromberg 1956. 53 Other organizations can of course be studied in a similar fashion, as providing a pure public good in a fashion other than militarily. See Olson and Zeckhauser 1966; Sandler and Hartley 1999. 52 51 21 object of the whole organization”, to be pursued through institutionalized collective action. 54 The military spending choices of the League members can be described with a rather simple Prisoner’s Dilemma arms race framework. Thus, the two participating nations in an international system or sub-system have two strategies: either to limit or escalate one’s military spending.
The hypothetical payoffs presented arise from four strategy combinations: 1) Both countries limit their ME; 2) Nation 1 limits while nation 2 escalates; 3) Nation 2 limits while nation 1 escalates; 4) Both countries escalate their military spending. The outcome shows that each nation is best off when it escalates and the other limits its military spending. The worst outcome for any nation, for example within the League of Nations, would be to limit military spending while another, especially a rival, escalates. 55 However, given the absence of trust or some type of guarantee, the escalate-escalate strategy would prevail, producing the worst outcome in terms of the maintenance of peace between the two nations.
In the 1920s, the League of Nations framework provided only a semblance of collective security, amply revealed in the 1930s. Although in general the 17-country (real) aggregate military spending remained quite stable after its sharp reduction in 1920??? 1922 until 1933, systemic stability was undermined by the decreasing spending shares of democracies. Also, whereas the eleven European states analyzed here maintained their military spending shares, on the average, throughout the 1920s and 1930s, their authoritarian challengers increased their relative spending strongly in the 1930s, thus tipping the balance between the democracies and autocracies after 1933.
Moreover, not all states were willing to disarm in the 1920s, given the League’s vague collective security enforcement guarantees. Only the naval disarmament seemed to produce more concrete results, although even this process produced results that were not entirely beneficial for the power balance. 56 Thus, the “disarmament equilibrium” achieved in the 1920s was a tenuous one. The dominant strategy of the League members, which emerged in the 1930s, was to escalate military spending again within the constraints 54 55 The League of nations starts; an outline by its organisers 1920, 26; Steiner 1993, 38. Sandler and Hartley 1995, 74-76, which also presents the ordinal version of this game. 22 placed upon them by their respective political economies.
The absence of successful repeated interactions that would foster trust between the key states (such as in achieving comprehensive disarmament measures) and the impotence of the central arbiters, the League and the United States,57 were among key reasons for the escalation of military spending in the 1930s. The failure of the League to include all the important world powers was of course in itself of paramount importance for its effectiveness. The isolationism of the United States; the distrust of the Soviet Union towards the League; Germany’s at first externally imposed exclusion in the 1920s and then its own decision to abandon the League in the mid-1930s; and Japan’s exit from the League on the heels of the Manchurian Incident were all huge blows to the collective peace aspirations.
After all, a key idea in the League structure was, contrary to the wishes of France and Eastern European states, that the enforcement of peace should be left to the hands of the members themselves. Especially the British were in favor of this interpretation of the League commitments. A comment attached to the presentation of the original League Covenant to the Parliament in June 1919 is quite illustrative: “It is true that, in default of a strong international striking force, ready for instant action in all parts of the world, the Members of the League must make their own arrangements for immediate self-defence against any force that could be suddenly concentrated against them, relying on such understandings as they have come to with their neighbours previously for this purpose. “58
The French, however, firmly believed that an international military force, preferably on permanent basis, or at least extensive military cooperation would be necessary in order for the League to function properly. The British and the Americans obviously did not want to make such binding arrangements. The British wanted to minimize their involvement in continental matters, especially in military affairs59. France was perhaps See Eloranta 2002b. On the comparison between the pre-war arbitration movement and the more extensive League of Nations, see Egerton 1974. 58 The League of nations starts; an outline by its organisers 1920, 234-235. 59 This was hardly a unanimous view held by British officials, especially those serving in the League of Nations. See Towle 1993. 57 56 23 he most ardent advocate of collective security guarantees among the Great Powers, yet it mainly viewed the League as a system of force directed against real or imagined German aggression. Their views on disarmament differed drastically as well, since for example the British were not willing to commit to extensive collective security arrangements, and the French were not willing, in the absence of such commitments, to disarm. Both were at best sceptical of the chances of the League to provide real security solutions. 60 While the accession of Germany to the League seemingly fostered the “spirit of Locarno”, in fact very little changed in terms of the League’s credibility in the 1920s.
Good diplomatic relations between the European “Big Three” (France, the United Kingdom, and Germany) did facilitate the everyday functions of the League, yet these superficial improvements were undermined by the representatives of their respective governments in their nationalistic domestic appearances. The “Big Three” only really turned to the League as an instrument of last resort. According to George Scott (1973), “Britain’s influence was often the most decisive in keeping the League out of things. “61 The role of the “weak” states, carrying the connotation of limited political and/or economic importance, was not quite as unimportant as their economic position in the international system would suggest. 62 This was guaranteed by the very structure of the League organs. Both the Assembly and the Council ensured the Small and Medium Powers63 practically equal say in the matters of the League.
Moreover, as Ronald Stromberg has pointed out, the “small” states were not necessarily any more virtuous than the Great Powers in international politics. 64 They also pursued their own interests and agendas within the League. There were also many types of “weak” states, even within Europe, representing a heterogeneous array of views. For example, the Scandinavian states ??? consisting of Denmark, Norway, and Sweden ??? felt that because of their geographic and political positions they were not willing to even entertain sanctions as a coercive tool, let alone military force. They attached, on the Northedge 1986, 44; Stromberg 1956, 252-253; Egerton 1974; Steiner 1993, 37. Scott 1973, 161-166. See also Towle 1993. 2 See Eloranta 2002a for further discussion. 63 See e. g. Ray and Singer 1973 and in general Singer 1981 for discussion on how to construct indices, based on diplomatic representation, to measure power status. See also the discussion in Eloranta 2002b. 64 Stromberg 1956, 263, footnote 14. 61 60 24 basis of their small populations and military weakness, very little value to armaments as the basis for security, and they were ready to implement general disarmament, even unilaterally, without further guarantees. The position adopted by these states maintained that general disarmament itself would constitute an important guarantee for international security. 65
There were also other constellations among the “weak” states that served as the basis for their activities in the League of Nations and beyond. For example, the so-called Oslo states ??? a group which came into existence in 1930 and lasted until 1940 ???consisted of Sweden, Norway, Denmark, the Netherlands, Belgium, Luxembourg, and Finland. The principal aim of this group was to promote economic and political cooperation, mainly under the auspices of the League of Nations. In reality, most of the cooperation took place directly between the Oslo states. However, since the core of this group consisted of the three neutral Scandinavian states, there were many disagreements over policies between them.
For example, whereas Belgium and the Netherlands were in favor of extending political cooperation to the so-called Group of Eight ??? which in addition to the Oslo states (not counting Luxembourg) comprised Czechoslovakia, Spain, and Switzerland ??? the Scandinavian states were in favor of forming an entity based around the former neutral states. Similar difficulties plagued the actions of the Group of Eight, which was mainly aimed at promoting the disarmament process within the League. Differences between the positions adopted by these states, for example in disarmament, were often irreconcilable. For example, Finland and Czechoslovakia were strongly in favor of security guarantees, whereas the Scandinavian states were willing to undertake unconditional, unilateral disarmament.
The Scandinavian policy of noninvolvement actually extended to denying automatic assistance to the victims of aggression, thus going against the proposal made by Finland in the League. 66 The ambiguous foreign policy of the southernmost member of the Group of Eight, Spain, was not particularly helpful for the cause of disarmament either. Primo de Rivera’s authoritarian regime, lacking fascist-style state control, maintained an 65 66 Jones 1939, 217-225; van Roon 1989. See also Salmon 1997. van Roon 1989, e. g. 128, Chapter IX; Jones 1939, 238; Salmon 1997. 25 ambiguous dualism in its foreign policy, alternating between a revisionist stance and a traditional policy of accepting the status quo. Rivera considered success abroad to be vital for the survival of his regime.
It maintained a lukewarm diplomatic courtship with Italy as an ongoing process while attempting to make headway in Northern Africa, mainly at the expense of the French. Despite some modest successes due mainly to luck, its biggest challenge was to obtain greater recognition from the League. Since the beginning, Spain’s seat in the Council had been a nonpermanent one, which was a source of irritation for the regime. As Germany acceded to the League in 1926 and obtained a permanent seat, Rivera began a campaign to get Spain a permanent seat as well, which after numerous twists and turns resulted in Spain’s withdrawal from the League for two years in the late 1920s.
When Spain rejoined, it did so under the same conditions as before, having achieved practically nothing except a political embarrassment and ending up undermining the League. 67 During the Second Republic, the foreign policy reversal of the regime was quite extensive. The Republican government relied on international cooperation and the promotion of pacifistic ideas. It was actually Spain’s initiative that led to the establishment of the Group of Eight. Spain demanded the greatest possible disarmament compatible with a guarantee of internal public order and the fulfillment of international obligations. Yet, although Ismael Saz claims otherwise, this group was far from uniform and imposed its own difficulties to the disarmament process.
Thus, it was a combination of competing visions among the Great Powers and the other states that ultimately made the disarmament compromise impossible. 68 Neutrality was not really a choice that Spain embraced willingly rather than a continuous descent towards “a hesitant and frequently shameful neutrality”, as the cooperation with the other “weak” states failed to produce results. 69 If we think of the disarmament as a game, the near impossibility of the disarmament becomes apparent. When the participants in the game had, broadly speaking, either the Saz 1999a, 53-64. See also Lee 1987, 227-231. Cf. Saz 1999b, 77-99. On competing visions of the reasons behind the disarmament process, see Vaisse 1993. 69 Saz 1999b, 83-91.
In addition, see Linz and Stepaan 1978. 68 67 26 goal of obtaining comprehensive collective security guarantees (like France) or, at the other end of the spectrum, were willing to accept disarmament without any agreement at all, the disarmament process certainly faced an uphill battle in order to be a success. At the level of individual countries’ foreign policy, the aims and motivations of the participants differed even more drastically. Furthermore, when repeated negotiations failed to produce results and centralized military leadership ??? either by the League of Nations or the leader nations ??? was not forthcoming, an arms race ensued in the 1930s.
Even the disarmament process of the 1920s was most likely a phenomenon that had more to do with individual state public finances and other domestic factors. The response by the autocracies, encouraged by their populist leaders, was to achieve revisionist aims (like Germany) or simply take advantage of the existing power vacuum. The response of the democracies was a more protracted one; i. e. , they retained the strategy of non-escalation until the mid- or late 1930s, which of course meant that the strategic payoff for the autocracies from the rapid arms buildup was even greater. It also is possible to test the notion whether the League actually produced a pure public good deterrence.
First, however, the representativeness of the sample must be discussed. The eleven European states, based on Eloranta (2002b), can hardly be said to represent the whole of the League of Nations, yet they could be argued to represent the European dimension of the organization quite well. Finland, Spain, and Portugal represented the periphery of Europe, the Scandinavian countries formed their own, fairly distinctive group, whereas the others could be viewed as belonging to the European core in terms of geography and level of economic development. Such countries as Germany and Japan, for example, did not belong to the organization during the whole period.
In this fashion, Spain perhaps is the least fitting of the group. Nonetheless, Spain’s position in the League or its military spending were hardly affected by its short absence. Moreover, France and the United Kingdom were most certainly the leading states in the League, so perhaps the representativeness of the sample of eleven is better than one might at first expect. Certainly this sample should reveal the effectiveness or inadequacy of the League of Nations as a provider of collective security. 27 Furthermore, it is possible to perform relatively simple tests to see whether military spending was a pure public good among the selected sample states.
If it actually was, then the League of Nations obviously was more important for these nations than the previous discussion indicates. Beginning with Mancur Olson and Richard Zeckhauser’s path breaking work on NATO, there have been many testable hypotheses relating to the idea of collective security provision in an organization and the implications of this provision on military spending. 70 Given the free riding tendency in a pure public good alliance, military burdens are anticipated to be shared unevenly in an alliance; thus, large wealthy allies (measured by real GDP) should shoulder more of the common defense than the smaller, poorer allies (=hypothesis 1).
The logic outlined by Olson and Zeckhauser maintained that a nation with a large area, long frontiers, and a higher population density, in addition to a higher share of vulnerable resources and ideological tendencies, would lead to a more aggressive military spending policy. 71 Of course, as I would argue, the explanation resting on the foundation of the public good theory and the suboptimality of defense provision via the spillover effect has sound theoretical foundations. 72 Indeed, Olson and Zeckhauser found a significant positive correlation, using Spearman rank correlation tests73, between the NATO allies’ GNP and their military burdens in 1964, indicating clear free-riding behavior by the smaller allies. 4 Later studies specified the pure public good alliance to describe the NATO until 1966, when the positive rank correlation between the variables ceased to be statistically significant. 75 Here I will perform the Spearman rank correlation tests between the military burdens and the real GDP levels76 among the selected eleven European states for three cross-section years (1925, 1930, and 1935). See Olson and Zeckhauser 1966 as well as Sandler and Hartley 1999 on NATO. Olson and Zeckhauser 1966, e. g. 371. 72 See Eloranta 2002. 73 Spearman R assumes that the variables under consideration were measured on at least an ordinal (rank order) scale, that is, that the individual observations can be ranked into two ordered series.
It is a nonparametric test, which is well suited for the analysis of small samples in particular. The null hypothesis is a zero coefficient. 74 See Olson and Zeckhauser 1966. 75 Sandler and Hartley 1999, 44 and the studies listed therein. Some years in the 1980s did produce a similar pure public good impact as the yearly years of the NATO. 76 Utilizing the modified GDP data used in most statistical exercises in Eloranta 2002b. 71 70 28 Respectively, defense spending should be allocated inefficiently from an alliance standpoint, as the sum of marginal benefits from defense provision should not equal the marginal costs of this provision (=hypothesis 2).
This follows from the argument that the military burdens in an alliance yielding joint products should be shared based on the benefits received ??? the greater the ratio of excludable benefits to total benefits, the larger should be the agreement between the benefits received and burdens shared (=hypothesis 3). For example Todd Sandler and Keith Hartley (1999) used the mean of three benefit shares (an ally’s share of NATO population, share of NATO’s GDP, and share of NATO’s exposed borders) to measure the sum of benefits, whereas the cost variable was the military burden. Using the Wilcoxon test77 that indicates whether the two measures are statistically the same, they found matching cost-benefit shares for most of the post-1966 period, indicating the presence of joint products (i. e. , impure public good benefits). 8 Here I will first perform the Spearman rank correlation tests between the real ME share and the ECONCINC (=arithmetic average of various economic indicators in a 17-country system), as explained in Eloranta (2002b), for the same three cross-section years as indicated above. Furthermore, I will perform the Wilcoxon tests to see whether the cross-section variables were statistically the same. Thus, if the results indicate the presence of joint products, we may deduce that a central authority in an alliance (here: the League of Nations or at least its European core) is required to coordinate spending to overcome the suboptimal provision and ensure the functionality of the cooperation (=hypothesis 4). As we have seen, no such leadership was forthcoming either from the League itself or outside this organization.
Relating to the arguments presented in the previous sections as well as the exploitation hypothesis, I will investigate whether the level of economic development, measured by real GDP per capita, might be an important explanatory variable in the burden sharing. Thus, we may hypothesize that the more developed the nation is econ