Introduction: Defence Alliances and the Economic Debate
In the years that followed the end of the Second World War, European governments started pursuing their national security objectives through membership of
military alliances in Western and Eastern Europe (Whitman, 2006:45). Throughout the Cold War, NATO and the Warsaw Pact were the frameworks for the provision of military security. With The end of the big conflict East-West symbolically market by the fall of the Berlin Wall in 1989, the dissolution of the Warsaw Pact was followed by doubts as to whether NATO would continue to have a raison d’être. More than 50 years have passed and NATO remains an important organization that frequently adapts its missions to account for new realities.
A large body of literature has examined the efficiency of defence alliances by stressing the public and private benefits of alliance security. It is frequently argued that defence burdens will be shared unevenly among allies and security provisions will stop before the optimal level because defence is a public good available to all, irrespective of whether everyone has contributed to its supply.
This essay will investigate the experience of NATO from an economic perspective in an attempt to assess the proposition that international defence alliances are inefficient because their goal is a public good. In other words, this essay is devoted to assess the efficiency of defence alliances with an eye towards shedding light on the actual relevance of economic interpretations.
It begins with a brief explanation of the economic theory of alliances elaborated by Olson and Zeckhauser in 1966 that characterizes defence as a pure public good. In a second moment this essay turns to the case of NATO. In the analysis it will become evident that defence provision also yields national or non-collective benefits. Some limitations of the economic debate will also be addressed. In the final part main conclusions will be highlighted.
From the economic standpoint, this essay will suggest that the proposition that international defence alliances are inefficient is not entirely correct. Alliances will work more efficiently whenever the gains from collective action have the characteristics of a private good. This essay also stresses that, despite all the difficulties associated with the collective provision of goods, international cooperation is likely to remain essential given the global nature of the new security challenges of the 21st century.
The theoretical frame of reference
The following paragraphs briefly present the economic theory of alliances first developed by Olson and Zeckhauser (1966) and later adapted to account for changing conditions of NATO.
The model outlined by Olson and Zeckhauser was an attempt to explain the functioning of NATO and other international organizations established by independent nations to further their common interests. In this model, defence is characterized as a pure public good because it satisfies two basic requirements: the benefits associated with defence are non-excludable and non-rival among allies (Sandler and Hartley, 1999a:29). The benefits are non-excludable because once security is provided all the members, including non-payer, automatically benefit (Sandler, 1993:448). In this case, due to collateral damages, the security provider will always have incentives to deliver the agreed retaliatory response against an invader of another ally (ibid.). In other words, “non-purchasers cannot feasibly be kept from consuming the good” (Olson and Zeckhauser, 1966:267).
The benefits related to security are non-rival or indivisible because one unit of defence can be consumed by one participant of the alliance without diminishing the consumption benefits from that same unit of defence still available to other participants (Sandler, 1993:448). There are no marginal costs associated with extending the benefits of a pure public good to another agent (Sandler and Hartley, 1999a:29).
If military alliances are qualified as sharing a purely public good then a number of implications can be conceived. First, alliance wide defence is characterized in the model as the simple sum of the allies defence activities (Sandler, 1993:449). This idea implies perfect substitutability of defence among allies in the sense that one extra unit of defence provided by any of the allies enhances alliance wide defence (ibib.).
Second, because all the members enjoy the benefits provided by the alliance regardless of whether or not a payment has been made and because defence provisions are perfectly substitutable inside the alliance, one might come to the logical conclusion that defence burdens will be shared disproportionately among allies (Sandler and Hartley, 1999a:30). Only those who value defence most will provide it whilst the rest has no incentives to contribute and will free ride on the provision efforts of others (ibid.). The model recognizes that a nation’s valuation of alliance forces varies among allies due to different factors, one of them being national income. The authors suggest that wealthier nations are likely to place higher absolute value on the alliance good due to the assumption that defence tends to be income normal (ibid.). If this is true, demand for defence is expected to increase with national income and the value derived from defence depends directly on the wealth of the nation (ibib.).
Third, and most importantly due to the purpose of this essay, defence expenditures in alliances are predicted to be at suboptimal or inefficient levels in relation to a Pareto-optimal standard (Sandler, 1993:446). The understanding of this specific point requires further explanations. According to the line of reasoning of Olson and Zeckhauser, each ally gets only a fraction of the benefits of the collective good provided even though each ally pays the full cost of any additional unit of security (Olson and Zeckhauser, 1966:278). As a consequence, rational individual members will rely on the provision of others and will have incentives to stop providing the collective good before the Pareto-optimal output has been provided (ibid.). Pareto optimality is a well-known concept of welfare economics used to describe a situation when it is no longer possible to reallocate resources in such a way that we can make one participant better off, without at the same time making another participant worse off (Griffiths and Wall, 2000:431). In the model analysed, defence spending is expected to be allocated inefficiently from an alliance standpoint because members will not reach a situation of Pareto-optimal output (Sandler and Hartley, 1999a:31). When defence is a purely public good, marginal costs are not shared in the same proportions as the benefits of additional units of the good. Because marginal costs associated to the provision of security are not equated to the sum of marginal benefits it is possible to say that alliances guide to a non efficient equilibrium (ibid.). When defence is a public good and there are no institutional arrangements or central authorities that alter the pattern of incentives, nations forming an alliance will fail to provide the level of forces that they have themselves recognized as appropriate in their common interest (Olson and Zeckhauser, 1966:267; Frey, 1984:128).
Forth, because defence benefits are purely public, an alliance can increase in number of members without subtracting from the amount of defence available to current members (Olson and Zeckhauser, 1966:273). In other words, there is no need to restrict the entrance of new members in the alliance.
This section presented the central economic theory applied to the study of international defence alliances known as the pure public good model. When the work was published, in 1966, the world faced a moment of particularly political relevance. In the 1960s, the protracted conflict and competition between the United Stated and the Soviet Union had reached its peak and a nuclear war seemed more feasible than ever. The confrontation regarding the Soviet deployment of nuclear missiles in Cuba in 1962, known as the Cuban Missile Crisis, was one of the main tensions that marked this period.
The following section will analyse the experience of NATO with an eye towards shedding light on the economic debate of this specific defence alliance. In the analysis it will become evident that alliances also provide national, non-collective benefits and this crucial fact alters the pattern of incentives, decreasing the importance of disproportion and the degree of sub optimality.
The case of NATO: Past and Present
After the publication of Olson and Zeckhauser’s article, many different scholars have applied the model on subsequent studies of NATO. This section will analyse the case of NATO from its origins to the present from an economic perspective. First, different phases or eras identified in the work of various academics will be presented in an attempt to demonstrate that alterations in strategic doctrine and weapon technologies can change the mix of public and private goods associated with defence activities. Secondly, this section will extend the analysis and consider the difficulties associated with NATO’s collective provision of security in the present and in the likely future.
Throughout the 1950s and the first half of the 1960s, NATO adopted a strategic doctrine of mutual assured destruction (MAD) implying that a nuclear attack would be initiated as a response to any Soviet territorial expansion involving NATO allies (Sandler and Murdoch, 2000:304). In comparison to NATO allies, which had converted a large share of their defence industries to peacetime uses since the end of the Second World War, the Soviet Union had acquired by 1950 an advantage when considering conventional military weapons (ibid.). The strategy adopted meant that NATO would rely on US strategic nuclear weapons to deter Soviet use of its conventional superior weapons in Western Europe (Sandler and Hartley, 1999a:37).
Due to the superiority of the US strategic nuclear forces, the American threat to retaliate on behalf of its European allies was credible and a large share of the defence output derived from NATO’s defence efforts was deterrence (Sandler and Hartley, 1999a:38). The benefits of deterrence, as provided by a nuclear triad, had the characteristics of a public good. In this specific phase of the Cold War, the retaliatory response was likely to be executed and the American ability to deter Soviet aggression was independent of the number of allies (Sandler, 1993:448).
Therefore, it is possible to say that, in the period considered, the economic theory of alliances developed by Olson and Zeckhauser that characterized defence as a pure public good seemed to apply to NATO. If this is true, then the implications of the model described in the previous section of this essay should also apply. Different studies have tested the correlation between the allies defence burdens and the GDP ranks for that period and the results were consistent with the pure public model (Sandler and Murdoch, 2000:308). Put simply, the US was carrying the defence burden of the other allies during the MAD era. Because NATO members were relying on the US provision of security, from the economic point of view it is possible to say that alliance resources were not being committed efficiently.
The situation changed in the second half of the 1960s with the adoption in 1967 of a different strategic doctrine that no longer relied on US nuclear force superiority to control the ambitions of the Soviet Union (Sandler and Hartley, 1999a:38). Under the new doctrine, the alliance would respond to one provocation with a measured response based on the nature and the seriousness of the aggression (ibid.). Therefore, the automaticly of nuclear retaliation was no longer credible and conventional weapons would be primarily used to respond to an invasion. It is also true that the Soviet Union had built up its strategic nuclear arsenal, weakening the US superiority and increasing the caution at both sides (ibid.:39).
The new doctrine of flexible response inserts new elements in our analysis and challenges the pure public good model that until this point seemed to apply. An examination of the percentages of incomes the members of NATO devoted to defence during the 1970s shows that the defence burden for the United States had dropped in comparison with the previous decade, while those of the other allies had stayed in the same level (Sandler and Hartley, 1999a:42). The numbers reflected the narrowing of the burden-sharing gap and indicated that for some reason the pattern of incentives had changed.
To explain the new reality the pure public good model was generalized to incorporate the fact that the benefits of alliance security also provide impurely public benefits (Sandler and Hartley, 1999a:34). The generalisation gave rise to the joint product model. When the benefits derived from defence activities are partially excludable by the providing ally or partially rival among the allies, then these benefits are impurely public (ibid.:34).
The alteration in the strategic doctrine, emphasizing the complementarities of nuclear, tactical and conventional weapons, affected the mix of excludable and non-excludable defence benefits shared by NATO allies (Sandler and Hartley, 1999a:47). Conventional weapons are damage-limiting and one or more allies can be excluded after deployment decisions (ibib.). In addition, they offer protection only to a limited geographical area, leaving other areas or borders unprotected and vulnerable to attacks (ibid.:35). Therefore, conventional forces in an alliance provide benefits that can be partially excludable and subjected to consumption rivalry, the two basic requirements of public goods.
In this new scenario, the idea of perfect substitutability of defence among NATO allies was no longer valid and countries had less incentive to free ride on the US provision of defence (Sandler and Murdoch, 2000:304). This is the main explanation behind the closing of the burden gap. From the economic standpoint, defence burdens started to be shared based on the benefits received (Sandler and Hartley, 1999a:34). This indicates that defence output is closer to the situation of Pareto-optimal and the alliance is moving towards a more efficient equilibrium.
The historical period of NATO that comprises the doctrine of flexible response, from the mid 1960s to the end of the Cold War in 1991, is extremely relevant in this study. It shows that defence activity can also give rise to impurely public and country-specific benefits. During this phase, it was no longer possible to rely on the military superiority and the automatic retaliatory response of the US to guarantee national protection. The sense of security became a product of domestic actions. The new scenario indicated that “allies who failed to maintain conventional forces became the weak link that might draw an attack” (Sandler and Murdoch, 2000:304). In this case, defence output was no longer essentially a public good. Based on this period and considering the economic approach applied throughout this essay, it is possible to say that international defence alliances can be more efficient and manage to provide levels of security close to the Pareto-optimal standard when private benefits are greater than public benefits.
Having presented the two different phases of NATO in the Cold War period, stressing distinct military doctrines and the way they affected the efficient allocation of alliance resources, this section now turns to the analysis of the recent period of this organization.
The end of the Cold War was marked by different developments that changed the nature of European defence. To account for the new reality, NATO had to adapt itself and its defence doctrine. Since the end of the conflict East-West, the alliance has incorporated new missions and it is likely that this process of adaptation will continue in future years.
Throughout the 1990s, the alliance reinforced its new commitments to peacekeeping, as part of its new strategic crisis management doctrine (Sandler and Murdoch, 2000:306-307). The new doctrine adopted once again affects the mix of public and private goods and the commitment of resources associated with defence activities.
The crisis-management shares of the allies defence budget were still small during the 1990s (Sandler and Murdoch, 2000:307). However, this movement is expected to become more evident in the 21st century and is presumed to increase pure public outputs (Sandler and Murdoch, 2000:307). The literature identifies at least three main reasons for that. First, when successful, peacekeeping activities promote world stability and overall security that benefit all nations including the ones who have not financially contributed (ibid.). Second, only the largest allies, and that includes the US, the UK, France and Germany, were investing in power-projecting capacity in order to be able to project their forces to areas of conflicts beyond NATO boundaries (ibid.). If only a limited number of allies are making the necessary investments, then the situation is likely to stimulate a free ride in times of crises for allies that have not made efforts to invest in this capability. Third, the necessity to be rapidly projected to new theatres calls for a next generation of weapons, highly associated with new technologies and R&D breakthroughs that can yield non-rival though excludable benefits (ibid.). This follows because deployment decisions can protect only limited areas, stressing non-rival benefits. The provider may not want to share the military technology, stressing excludable benefits, but he will share it whenever he understands that this attitude increases his level of security.
Focusing on the percentages of incomes the members of NATO devoted to defence during the 1990s, researchers found no evidence of disproportionate burden sharing (Sandler and Murdoch, 2000:311). However, they found a small increase in the rank correlation in the final years of the decade, leaving an indication of increased disproportion in the future (ibid.).
The focus of this essay being on the economic debate of NATO, it is essential to examine the recent numbers that reflect defence expenditures as a percent of gross domestic product. Observing the figures until 2005 , one might come to the conclusion that there are indications of the consistency of the pure public good model. In other words, it seems that the defence burden for the United States increased in the period 2000-2005, while it dropped or remained relatively constant for the other allies. If this interpretation is correct, then it is possible to say, from an economic perspective, that resources are not being efficiently committed by NATO allies nowadays. In other words, small members are once again free riding on the US provision of defence and, as a result, the overall level of defence provided is inferior to the suboptimal level. NATO’s increased commitment to the new crisis management doctrine is one possible explanation for that, as explained above.
So far, this section has analysed the case of NATO from its origins to the present from an economic standpoint. In the analysis, some of the difficulties associated with the collective provision of security became evident. The proposition that international defence alliances will always be inefficient because their goal is a public good does not seem entirely correct. It has been demonstrated that military expenditures are a common public good to a limited extend because they also serve purely national purposes (Frey, 1984:132). NATO’s strategic doctrine, weapon technologies, perceived threats and other factors, can influence the mix of public and private goods associated with defence activities. Nevertheless, from the economic perspective, it can be argued that alliances will not work efficiently whenever there is a high degree of publicity associated with the benefits of defence provision. This situation generates disappointments and protests within the alliance, since the burdens of the common defence are being shared disproportionately (Olson and Zeckhauser, 1966:266). It also raises concerns about the level of security provided, which is below the level nations themselves have defined as optimal or necessary for an efficient military action (ibid.).
In the future, it is possible that NATO’s strategic doctrine will have to be adapted to incorporate new missions and threats. Transnational terrorism and environmental degradation are some of the challenges that NATO will most likely have to address in the years to come (Sandler and Hartley, 1999a:166). Given the global nature of the new security challenges, the benefits associated to defence expenditures might have the properties of a public good and, in this case, nations will have more incentives to free ride on the provision of others (ibid.:178). If NATO redesigns its missions to account for these new global challenges, then it might be suggested that the US will continue to bear a disproportionately large share of the common burden. From the economic standpoint, the alliance will not commit resources efficiently and nations will have incentives to stop providing the good before the Pareto-optimal standard is reached.
It is important to stress that the economic evaluation of international defence alliances applied throughout this essay have limitations. The reality is often more complex and political reasons are always taken into consideration by the member states when they have to decide how much money will they devote to security. The impact of interest groups linked to the military-industrial sector and the influence of bureaucrats are some of the factors that can influence the provision of security. If this is true, then it is possible to say that nations do not always act like rational individuals who try to maximize their utilities.
The fact that there are political elements involved also explains why, despite regular claims about unfair burden sharing and inefficient level of forces provided, no ally has ever withdrawn from NATO (Sandler and Hartley, 1999b:676). Nations remain members of the organization not only because they understand it offers more protection or lower defence costs than they had before, but also because there are political reasons to be considered. The US might place a sufficiently high value on world leadership and NATO participation and this compensates the fact that it devotes larger percentages of its national income to defence. Different political elements involved in international defence alliances help to keep members attached to the organization.
In addition, it can be argued that given the global nature of the new security challenges of the 21st century, NATO allies will have to cooperate with respect to deterrence decisions and engage in a greater degree of cooperation than in the past (Sandler and Hartley, 1999a:200). The nature of international terrorism, in particular, has reached such a complex stage outside of the scope of a state-centric analysis. To provide one example, secure US borders drive terrorists to attack US interests abroad (ibid.:173). As a consequence, “the US must rely, in part, on other countries to protect its interests against terrorism” (ibid.). Based on that, one might argue that NATO and other defence alliances are likely to remain significant in the international scenario and the US is expected to remain committed to European security despite long-standing debates about burden-sharing.
From the economic standpoint, this essay has analysed the proposition that international defence alliances work inefficiently because their goal is a public good. With reference to experience in NATO, it has stressed how alterations in strategic doctrine and weapon technologies have changed the mix of public and private goods associated with defence activities over the years.
It has been shown that alliances will not work efficiently whenever the degree of publicity associated with the benefits of defence provision is high. In this case, institutional arrangements or constitutional agreements could be applied to alter the pattern of incentives and guide to a more efficient equilibrium. Nevertheless, this essay has stressed that when the gains from collective action have the characteristics of a private good burdens of defence will be shared more in accordance with the benefits, leading to a situation of economic efficiency.
Despite the difficulties associated with the collective provision of security, it has been stressed that political elements also alter the pattern of incentives and influence the provision of security. Finally, this essay suggested that due to the global nature of new security challenges, international cooperation is likely to remain essential.
CHALMERS, M., 2001. The Atlantic Burden Sharing Debate: Widening or Fragmenting? International Affairs, 77(3), pp.569-585.
FREY, B.S., 1984. International Political Economics. Oxford: Blackwell.
GATES, W.R. AND TERASAWA, K., 2003. Reconsidering publicity in alliance defence expenditures: NATO expansion and burden sharing. Defence and Peace Economics, 14(5), pp.369-383.
GRIFFITHS, A. AND WALL, S., 2000. Intermediate Microeconomics: Theory and Applications. 2nd ed. Essex: Pearson Education Limited.
LUNN, S., 1983. Burden-Sharing in NATO. Worcester: Billing & Son Ltd.
OLSON, M. AND ZECKHAUSER, R., 1966. An Economic Theory of Alliances. Review of Economics and Statistics, 48(3), pp.266-279.
SANDLER, T., 1993. The Economic Theory of Alliances: A Survey. Journal of Conflict Resolution, 37(3), pp.446-483.
SANDLER, T. AND HARTLEY, K., 1999a. The Political Economy of NATO. Cambridge: Cambridge University Press.
SANDLER, T. AND HARTLEY, K., 1999b. NATO Burden-Sharing: Past and Future. Journal of Peace Research, 36(6), pp.665-680.
SANDLER, T. AND MURDOCH, J.C., 2000. On Sharing NATO Defence Burdens in the 1990s and Beyond. Fiscal Studies, 21(3), pp.297-327.
SLOAN, S.R., 2003. NATO, the European Union and the Atlantic Community. Lanham: Rowman & Littlefield Publishers.
VOIGT, K.D., 2004. Dealing with Terrorism: The EU and NATO. In: H. GARDNER. NATO and the European Union: New World, New Europe, New Threats. Hants: Ashgate, pp. 171-180.
WHITMAN, R.G. Nato, the EU and ESDP: An Emerging Division of Labour? In: M.A. SMITH. Where is NATO Going? Oxon: Routledge, pp.44-65.