IRDISP-RESEARCH INSTITUTE FOR DISARMAMENT, DEVELOPMENT AND PEACE

ABSTRACT: Fine that there is the arms race, but what has Italy to do with it? Aren't the U.S. and the Soviet Union the promoters of such race? It is true that the two superpowers are the chief responsibles for the arms race. The chief ones but not the only ones. Italy has its share of responsibility as well. A smaller share, but not a negligible one. In absolute numbers, Italy's military expenditure in 1985 was the eighth of the world. As regards the number of men at arms, it is among the first fifteen countries. And the Italians are in the first six positions among the world exporters of armaments. The weight of the military sector on the whole of the Italian economy is still rather limited: The expenditure accounts for 2.7% of the gross domestic product; arms represent 2.7% of the wealth produced by the industry, and 2.3% of the exports. Moreover, the military threats to the security of Italy are less serious than those which many other international actors have to face - including many of our allies. Ther

efore, Italy is in a situation that offers many opportunities to contain the expenditure, experiment conversions to civilian uses of the military productions, and promote a realistic security policy aimed to achieve detente.

Unfortunately, these opportunities are dropped. In fact, there has been a tendency to expansion over the last decade which must be urgently stopped. As of the mid-seventies, Italy has become one of the major exporters of weapons systems, and its military expenditure exceeds the annual growth rates decided at the NATO level. That same period marks the rise of the supporters of a "new military role" for Italy in the Mediterranean. The "White Paper" presented by the Minister of Defense Spadolini in winter 84-85 summarizes and pinpoints these developments, obviously from the point of view of a person who supports them and hopes that they will continue. This book on the contrary highlights the doubts, the questions, the alternative proposals compared to what has been to this moment a monologue carried out by the establishment.

("ITALY AND THE ARMS RACE" - A counter-White Paper of defense - edited by Marco De Andreis and Paolo Miggiano - Preface by Roberto Cicciomessere - Franco Angeli Libri, 1987, Milan)

3. PUBLIC GOODS, FREE RIDING AND BURDEN SHARING IN NATO

by Mauro Marč

1. Introduction

Relations among states on the international level are normally in the form of cooperation and/or conflict. In practice, real situations are somewhere between these two extremes: more or less accentuated situations of conflict frequently arise within stable periods of cooperation, and vice versa. One example of this is the functioning and operation of many of the most important international organizations, such as the IMF, Gatt, etc.

It is small wonder, therefore, that the issues of cooperation, its evolution and prospects and the relations of cooperation and conflict among the various countries have long since been the object of research on international relations. A specific case of cooperation of some substance is that which concerns, for example, the supply and provision of collective goods and services on the part of certain states.

Clearly, such matter can be investigated from different points of view and can be exemplified by extremely different cases. The collective supply of certain goods and services on the part of several states can bear on the monetary field, the political and economic one as well as on the field of military security.

In this chapter we will investigate the problems relative to the collective supply on the international level (i.e. a joint offer by several countries) of a particular type of good: security, and in particular the security provided by Nato to its member states.

In view of the peculiar features of this good, the specialized literature normally resorts to the traditional theory of public goods to analyse the supply of this good and the conditions for its efficiency and optimality. In past years, the prevalent approach has been that of acknowledging the status of international public good to the good of "security", the supply of which presents problems and difficulties that are not easy to solve.

An approach which relies on the theory of public goods makes it possible to carry out an in-depth analysis of the goods supplied by international organizations (particularly defense), their characteristics and conditions of supply, the division of the costs linked to this supply. At the same time, it enables to introduce simplifying devices in the study of the mechanisms and incentives behind many countries' drive to join others in international organizations. Throughout this chapter we will refer to: a) in the first paragraph, to the theory of public goods, namely their characteristics of "indivisibility" and "external effects"; b) in the second paragraph we will endeavour to formulate a more detailed and realistic definition of the international public good of "collective defense" on the basis of the most recent findings of the specialized literature; c) in the third paragraph we will briefly discuss the hypothesis of free riding (1), the theoretical and empirical consequences of this hypothesis on the fun

ctioning of international bodies (namely Nato) and its strategic implications; d) in the fourth chapter we will analyse the key subject of this chapter, which is of great strategic importance: the problem of burden sharing, i.e. the division of the burden of defense among the various Nato member states; e) lastly, a few conclusive remarks.

2. Public goods

For the purposes of this discussion, we will briefly summarize the definition and characteristics of a public good. The American economist P. Samuelson has provided a fairly explicit definition of public good. According to this definition, collective consumption goods (as he calls them) are goods which everyone enjoys collectively; the consumption of these goods by each individual implies no reduction in consumption by any other individual (2).

John Head has instead singled out two main characteristics of public goods (3).

The first has been referred to as "indivisibility" or "joint supply". On other words, once a good with public characteristics has been supplied to one individual, equivalent quantities of this good with the same characteristics can be made available to other individuals at no extra cost. A public good is a joint supply good in the sense that once it is produced, any given quantity of the public good is available to all in the same degree. The supply of a given quantity of public good to an individual, and the supply of the same quantity to other individuals, are obviously joint products (4).

Such a rigid definition of public good, i.e. the hypothesis of the existence of a "pure" public good, inevitably raises problems. A closer investigation reveals that the most frequently used examples of public goods (the light provided by a lighthouse to ships, defense, television programs, clean air, freeways, etc) do not seem to match the characteristics of the definition. "Capacity limits" normally arise far before the good is made available to all to the same degree, and often there are variations in quality even before encountering capacity limits strictly speaking. For example, in the case of a jammed road, it is usually acknowledged that the service supplied in such circumstance is qualitatively inferior to the one supplied by the same goods when they are less intensely used (5). These considerations lead us to the need to adopt a more flexible definition of public good; also, the concept of pure public good should be more appropriately considered an "extreme case" (6). While on the one hand the "join

t supply" remains a key feature of public goods, on the other hand beyond a certain point, the additional consumption of one person implies the need for a corresponding reduction in other people's consumption (7).

The second aspect that characterizes a public good is its capacity, once it is supplied, to generate external effects, i.e. the property of "non-exclusion". A public good is a good for which it is not possible to restrict the benefits that ensue from its supply only to those who are willing to pay for this good (8). The external effects of a public good consist in the fact that its benefits become available to all once this good has been supplied. These benefits cannot be divided among the individuals according to their willingness to contribute for the public good, because once a similar good has been supplied collectively, no individual can easily be excluded from its consumption, except at a high cost - the supply cost of a given quantity of public good to an additional individual equals zero (9). As consumers of public goods, each of us is rival to none: since that which is available to a given individual is available to all, the individual consumption of a public good in no way reduces the quantity that

may be consumed by other individuals (10).

For external effects as for indivisibility, there do not appear to be goods with such rigid characteristics of external economies, i.e. goods for which it is appropriate to speak of "complete impossibility' of exclusion" (11) - the impossibility of excluding other individuals from at least part of the benefits. Head argued instead that there are several goods that present external economies of some substance without nearing a complete impossibility of exclusion.

The case of a "pure" public good should therefore be seen as an extreme case, with regard to which we can find the degree of "collectivity" of each good.

In real life, there are several cases of public goods: from public goods that are normally considered pure, such as TV or radio programs broadcast by a station, information and national defense (12), and goods with less evident features of "purity" such as, for example, a park or jammed road, unto goods that according to some are somewhere between public and private goods, with substantial characteristics of external effects and indivisibility, such as public health and education.

On the international level, moreover, there are various forms of public goods that are normally produced and supplied by international organizations: these range from the goods produced by a military alliance among states, such as Nato - the collective defense and security of the members of the Alliance, the safeguard of common resources (seas, oceans, regulation of whaling, exploitation of oil reserves in the oceans, etc), to the case of an international monetary system (or even currency agreements and monetary "regimes") (13). These international public goods are mostly the result of a cooperation among certain states, which are frequently associated into international bodies, which have a common interest in the production and supply of these goods.

3. The international public good of "defense"

Defense has been traditionally used to represent a pure public good, with its properties of non-competition and non-excludability (14). Rather in the same way as in a state, an organization of allied states for collective defense produces a public good, except that the member states of the organization replace the citizens of the state. Public goods are the typical products of the activity of the governments and of the international institutions. Olson's and Zeckhauser's model of the economic theory of alliance aimed to explain the tendency in the smaller and poorest Nato countries to behave like "free riders" - a problem which is still at the core of the controversy of the so-called "problem of burden sharing" (see following paragraphs). The conclusions reached by Olson and Zeckhauser (15) were that, unlike a situation in which the group interested in the supply of a public good is very large and in which the share of benefits enjoyed by each individual is very small--where, in other words, no one will have

an incentive to supply the good, and where behaviours of total free riding will probably prevail--in an organization that represents a well-defined number of states, these may have an incentive to supply the public good and to bear the costs related to its supply, but will tend to supply a suboptimal quantity of this good. Moreover, it is most likely that the larger members will tend to bear an overdue share of the supply costs of the public good.

In any case, it seems difficult to accept Olson's and Zeckhauser's model as valid. On the one hand, the theoretical premises of their model have significantly changed: the Atlantic Alliance no longer produces "solely" a pure public good--the deterrence provided by nuclear missiles--but also, jointly, other goods with private characteristics or with "impure" public characteristics. On the other hand, much caution should be used in drawing final conclusions regarding the contributions of the various countries in the costs of the Atlantic Alliance. As we will see later on, the indicators that gauge such contributions are scarcely reliable and liable to substantial criticism.

The traditional concept of defense as a pure public good, moreover, has raised in certain cases fiery discussions and a few attempts to qualify the characteristics of this good in a more defined manner. It has been said that despite the fact that the idea of pure public good has been valid to some extent in theoretical discussions, in the case of defense it is characterized by certain well-defined limits and descriptive shortcomings.

In recently published papers, Todd Sandler and some of his colleagues (16) have developed a different approach, aiming to investigate the nature of the "impure" public good (17) of defense. Sandler advanced the thesis that the changes occurred in the early '70s in Nato's military strategy, the development of new weapons systems and other factors yet have brought about changes in the product of the Western military Alliance, making it difficult to apply the model of pure public goods as proposed and used by Olson and Zeckhauser in the study of military alliances.

The chief changes in Nato's military strategy Sandler refers to are: the adoption of the flexible response, the installation of tactical nuclear weapons and SALT (Strategic Arms Limitation Talks). The joint production model developed by Sandler (18) aims to investigate the aspects of "deterrence" (i.e. a pure public good) and "protection" (in the sense of damage limitation, which has both private and public characteristics) of defense, underlining its characteristics as an "impure" public good.

The collective defense supplied by a military alliance is a mixed good (or an impure public good) which presents aspects both of a private and of a public good; it is somewhere between the extremes of purely private goods and purely public goods (19).

According to Sandler, as with the study of private goods, joint products are of great importance for public goods. A joint production is when the same intermediate input(s) produce(s) two or more products: the typical example is oil, which is used to produce fuel, kerosene and synthetic fabrics. Likewise, in the case of public goods, an intermediate input could produce one or more public goods, private goods or impure public goods (20).

The thesis advanced by Sandler is that the goods of defense fall into three categories: "deterrent, defensive and mixed". A defense good is exclusively of dissuasion, a "deterrent" defense good has the sole purpose of notifying a credible retaliation threat by the military alliance; it is exclusively "defensive" when it is used to "protect, delay or prevent" an attack; lastly, it is a mixed good when it includes, to varying degrees, both deterrent and defensive elements (21).

The defense expenditure of a military alliance such as Nato produces "deterrence" (a pure public good) - e.g. intercontinental missiles; "domestic protection" (an impure public good, because it presents some of the characteristics of a private good), whose function is the safeguard and defense from an attack - e.g. an infantry regiment to protect a boundary, and, more in general, conventional forces; also, "entirely private benefits" such as those produced by the use of armed forces for the maintenance of the national order, or the economic and industrial repercussions of the military sector on the civilian one, as well as domestic political benefits (22).

However, going beyond theoretical models, in concrete terms there has never been a univocal acceptance within the Atlantic Alliance of the "pure" public feature of the nuclear deterrence provided almost exclusively by the United States. One of the main motivations behind the decision of France--and Great Britain--to acquire their own nuclear deterrent, the "force de frappe", is one based on the perception of an insufficiency of the American deterrent for the purposes of an effective defense of the European--and therefore French--territory. An incorrect use of this deterrent or--even more--the fear of its "non-use" by the U.S. in the event of an invasion of Western Europe by the troops of the Warsaw Pact to avert a nuclear escalation, are the chief fears behind France's decision to become "autonomous" on the nuclear level.

4. The problem of free riding

The use of the theory of public goods in the study of military alliances calls for a discussion on a hypothesis of particular strategic importance, which is usually related to his theory, i.e. the hypothesis of free riding.

The problem is raised by referring to the above-discussed properties of public goods: indivisibility and external effects in consumption; the imperfect capability of the producer of the public good to exclude certain individuals from the consumption of this good (23).

A free rider is an individual who tries to benefit (or rather, benefits) from a public good without contributing (or contributing less than it should) to the production costs of this good (24). In fact, almost any paper contains at least one mention to the problem of free riding. Throughout past years, moreover, the theory of public goods has focussed more and more on the problem of correctly poinpointing the individual preferences (of individuals, groups and nations) concerning public goods and their supply.

The difficulties for the theory of public goods which the hypothesis of free riding produces are easily explained: assuming that each member of a society knows that other members are interested, like he/she is, in the supply and consumption of a public good (and provided no one can be excluded from benefiting from the public good once this has been supplied), there follows that each individual can hope to become a free rider. In other words, each person can hope to benefit from the public good without bearing the costs connected to its supply (25). However, a closer investigation of the problem reveals that such strategy is excellent individually, but not in a collective framework of mutuality: if all individuals choose to be free riders, there will be neither production nor supply of the public good at hand. Buchanan, on the other hand, has underlined that while recognizing that "a behaviour of this kind on the part of everyone would produce undesirable effects--in other words, there would be benefits for n

one--it is necessary to acknowledge that "the individual has no interest in participating of his own volition to the agreement, considering that he can obtain the desired effect while leaving others to produce the public good in the greatest possible quantity, whereas he would enjoy the advantages of being a free rider. In other words, he obtains the benefits without participating in the costs. Even if he were to sign an agreement on the participation in the costs, he would have a strong incentive to break the contract and cheat on the terms" (26).

While partly recognizing the validity of the hypotheses outlined by Olson and Zeckhauser, we believe that the problem of free riding is "essentially" linked to the problem of the size of the group of individuals or states for which the benefits cannot be divided (as Olson and Zeckhauser themselves admitted).

It has been remarked that the larger the group, the more relevant a potential free riding behaviour, hence the likelihood that the public good cannot be financed by voluntary contributions (27); in other terms, the wider the group, the more likely free riding behaviours. Therefore, the public good could ultimately cease to be produced or supplied. More practically, the problem could become relevant in a situation with "n" individuals or states (where "n" is a very high number) in which it is difficult to grasp the "real" individual preferences and establish the costs for the supply of the public good accordingly.

It seems to us that the hypothesis of free riding, in situations with a limited number of actors interested in the supply of the public good (such as Nato), is secondary, or at least not as important as traditionally provided by the theory.

A person who really acted as a free rider would ultimately be found out, and this would lead to threats and retaliation on the part of the other members of the Alliance and possibly even to a series of chain reactions, the outcome of which is hard to predict.

The behaviour of a state and its effects on the Alliance will be investigated by the other members and will influence their future behaviour (28).

It should be remarked that the existence of threats and retaliations, such as those in Nato, do not, per se, point to the existence of free riding behaviours. Even the "perception", based on subjective evaluations, on the part of one country of free riding attitudes by other countries is enough to trigger a "war" of threats and counterthreats and actual retaliation.

5. Burden sharing among Nato member states.

The economic theory of alliances, namely the hypotheses exposed by Olson and Zeckhauser, of a different contribution of the various countries to the common costs of the Alliance, leads us directly to the problem of dividing the common burden of Nato among the member states.

Over the past years, the question of burden sharing has been constantly on Nato's agenda. More or less frequently, the question of how the burden of resources necessary to prepare and ensure the safety of Nato should be calculated and distributed fairly among the various members has given rise to lively discussions and fiery contrasts among the allies (29). This problem has been especially important in the United States, since the Americans are convinced that they are bearing and financing an overdue share of Nato's common burden.

In the first years of Atlantic relations, the United States found it normal to bear a large part of the burden of Western security. It was convinced that the achievement of a safe Europe, both on the military and on the economic level, was in their best interest, and would have resulted in an advantage for it. Europe would have supplied an essential market for the American economy, which was expanding and needed to find outlet markets. However, as soon as Europe became richer economically, and therefore more independent (30), the Americans started to harbour the suspicion that Europe was behaving like a free rider in financing the Alliance's defense (31).

The core of the debate on the division of Nato's defense burden and of why this problem is mainly an "American concern", is the Americans' perception that the U.S. is providing to the defense of Europe far more than the Europeans themselves (329.

More recently, and in particular with the Reagan administration, the Americans have rekindled this controversy (33). L. J. Sullivan, a security advisor in Washington, has gone so far as maintaining (34) that no problem divides the Atlantic Alliance more than the question of burden sharing. According to figures provided by Sullivan himself, 57 per cent of the expenses for the defense of the U.S.--or 4 per cent of the GDP and 17 per cent of total US federal spending--are relative to the collective defense of Nato. On the basis of this, he believes that the agreements of the Atlantic Alliance can no longer be kept into effect, and deems it urgent to open a debate in the West on the division of strategic responsibilities.

On the other hand, the entire controversial question of the Nunn-Roth Amendment taken into examination by the U.S. Senate in June 1984, has clearly revealed the state of mind of the Americans and of a vast part of Congress. The Nunn-Roth Amendment (after its proponents) proposed a reduction of the U.S. troops destined to Nato by 30,000 units per year, for a total of three years, starting from 1987; this would have reduced the number of U.S. troops stationed in Europe by 90,000 units, from the current level of 326,000.

These cuts could be avoided only if the Europeans respected their commitment taken at Nato level (1978) to increase their defense expenditure in real terms at the annual rate of 3 per cent.

An amendment with more or less the same purposes, but lacking the mechanism of automatic cutbacks contained in the Nunn-Roth Amendment--therefore without the most incisive part for the allies--was passed (the Cohen amendment after the republican senator from Maine who introduced it) in that period. It contained the resolution that the U.S. would not have increased the number of its military personnel for Nato after 1985. In spite of the fact that the Nunn-Roth was rejected by the Senate (with a slight margin, 55-41), it did contribute to enhancing the Americans' already deep-seated belief that the European partners were not contributing enough to the common effort to defend the Atlantic Alliance.

In brief, the message to the Europeans was very clear: if the European Nato countries do not do enough to defend themselves (i.e. adequately contributing to the common expenses), why should the United States do it? (35). In spite of the fact that the question of burden sharing, as we will see later on, is essentially treated in an insufficient and simplistic way, and that the Europeans' contribution to the defense of the Alliance is largely underestimated, a growing number of Americans believe that Europe does not want to defend itself.

If we now analyse the question of burden sharing in detail, we must first of all discuss the problem of "how the burden is defined", and then analyse the problem of gauging and comparing the various countries' defense expenditure.

The use of an expression such as "burden sharing" necessarily involves the need for a consent on what the burden effectively includes; this consent is very rare (36). A definition of burden in terms of "resources necessary to guarantee the safety" of the Nato member states will inevitably imply differences in opinion among the various members on the meaning of the "safety" of a nation and consequently on what the burden implies.

The main reason for these differences lies in the different "points of view" expressed by the allied countries regarding the threat for the Alliance, and in the differences in opinion on what it is fair and necessary to do for national defense (37).

If we analyse, instead, the estimation of the defense effort of the allied countries, we can proceed by considering both the "contribution", i.e. the quantity of resources each country allocates for defense--in other words, refer to the "inputs" and their "cost"--and the "payoff", i.e. the actual military capacity activated by the defense expenditure, in other words the "output" of these expenses (38).

The evaluation of the input, i.e. of the resources allocated for defense, to compare the expenses of the various Nato countries, can be made by using several economic indicators (39). The main ones are: a) the total amount of the defense expenditure; b) the rate of real annual increase of the defense expenditure; c) the percentage of GDP (or GNP) allocated for defense; d) the per capita defense expenditure; e) the defense expenditure as percentage of total national expenditure; f) and lastly, of the other indicators aiming to gauge both the wealth (i.e. each country's potential contributive capacity) and the military expenditure of the various countries.

Thus, if we consider the defense expenditure of the various Nato countries expressed as percentage of the GDP (table 1), we can notice a consistent contribution on the part of the United States (second only to Greece), of Turkey and Great Britain, which all present percentages that are almost always above 5 per cent; there follows a group of countries (France, Portugal, Federal Germany, Belgium, Netherlands and Norway) characterized by a share of defense expenditure as percentage of the GDP ranging between 3 and 4 per cent; the last four countries (Italy, Denmark, Canada and Luxembourg) present the lowest shares.

-----

Tab. 1 Defense expenditure of the various Nato countries as percentage of the GDP

country 1980 1981 1982 1983 1984 1985*

Belgium 3,4 3,5 3,4 3,3 3,2 3,3

Denmark 2,4 2,5 2,5 2,4 2,3 2,3

France 4,0 4,2 4,1 4,2 4,1 4,1

West Germany 3,3 3,4 3,4 3,4 3,3 3,3

Great Britain 5,0 4,8 5,0 5,3 5,3 5,4

Greece 5,7 7,0 6,9 6,3 7,2 7,1

Italy 2,4 2,5 2,6 2,7 2,7 2,7

Luxembourg 1,2 1,2 1,2 1,2 1,2 1,2

Norway 2,9 2,9 3,0 3,1 2,8 3,2

Netherlands 3,1 3,2 3,2 3,2 3,2 3,1

Portugal 3,5 3,5 3,4 3,4 3,3 3,2

Turkey 4,3 4,9 5,2 4,8 4,4 4,4

Nato Europe 3,7 3,8 3,8 3,8 3,8 3,8

Canada 1,8 1,8 2,1 2,0 2,2 2,2

United States 5,5 5,8 6,4 6,6 6,5 6,9

Total Nato 4,4 4,7 5,1 5,3 5,3 5,6

* = Temporary estimates. Moreover, for Germany the figures have been reckoned without including the expenses for West Berlin; furthermore, for France the figures are merely indicative, as this country is not part of Nato's integrated military structure.

Source: "Nato's Sixteen Nations", Dec. 1985-Jan. 1986, pp. 86-87.

-----

Such figures and the ensuing overview - as the one resulting from table 2 (per capita defense expenditure) that highlights a strong contribution of the United States but, for example, a more modest role for Greece, Portugal and especially Turkey, and an increase in the importance of Norway - should be taken with extreme caution considering the substantial differences in the defense systems of the various Nato countries.

-----

Table 2 Per capita defense expenditure of the various Nato countries, in US $

country 1980 1981 1982 1983 1984 1985*

Belgium 270 273 272 272 273 289

Denmark 265 267 266 269 266 269

France 375 388 390 396 393 390

West Germany 358 369 367 371 371 379

Great Britain 396 402 422 429 442 452

Greece 205 249 245 223 260 258

Italy 160 159 163 167 171 174

Luxembourg 100 105 109 112 113 112

Norway 374 383 397 411 391 452

Netherlands 281 291 296 296 305 305

Portugal 71 71 71 69 65 65

Turkey 54 54 55 52 50 51

Nato Europe 270 276 280 282 285 289

Canada 235 240 248 263 281 285

United States 786 820 875 926 970 1040

Total Nato 473 491 515 538 558 589

* = Temporary estimates. Moreover, for Germany the figures have been reckoned without including the expenses for West Berlin; furthermore, for France the figures are merely indicative as this country is not part of Nato's integrated military structure.

Source: "Nato's Sixteen Nations", dic. 1985-gen. 1986, pp. 86-87.

-----

Considering the difficulty in disposing of effective output measures, the output (or payoff of the defense expenses) is normally calculated by analysing the way in which the expenses for defense are used, e.g. the percentage of resources in terms of personnel and equipment allocated for certain "key" areas, or the allocation within the defense expenditure for investments in equipment and machinery or also in training for the army, instead of less important sectors in terms of fighting effectiveness such as pensions (40). Both types of indicator - of the input and of the output - present considerable difficulties both conceptually, in their definition, and empirically, which make it difficult to analyse burden sharing in Nato.

Considerable problems are created by the fluctuation in the exchange rates of the various countries' currencies, as well as by the difference in the quality and in the indexes used to purge the data from inflation (41). A negative trend of the dollar on the international markets, for example, can make the Europeans' expenditure seem higher than it actually is, and vice versa (therefore, considering the dollar's trend over the last years, the Europeans' expenses in 1984 and 1985 dollars are underestimated).

As far as inflation is concerned, Nato has tried to make up for these difficulties by adopting an agreed-upon definition of the defense expenditure of the various allied countries, but has not yet managed to establish a homogeneous deflator capable of removing the effects of inflation from the defense expenditure of the member states.

Moreover, when using the indexes that compare the costs for personnel and, even more, the expenses for defense as percentage of the GDP, notice should be taken of the fact that the Nato countries have different recruitment and personnel management systems; in the countries with a compulsory military service, the lower salaries paid to the draftees keep the defense expenditure per capita and as percentage of the GDP at low levels, and do not enable a detailed and realistic comparison with the other countries. According to a certain number of estimates, if the Europeans' expenditures were adjusted according to the different salaries paid to the personnel, these would increase by 20 per cent, reaching an amount corresponding to that of the United States (42).

Other estimates have shown the influence of the size and different forms of the respective budgets on the indicators that show the percentage of national budget allocated for defense - more or less restrictive definitions of national budget; different tax systems.

Lastly, the fact that the indicators underline the trend of the defense expenditure of the various countries as well as providing information relative to each year is equally important (43).

Therefore, the conclusion seems to be that no single measure (the indexes relative to burden sharing among the Nato countries) provides a valid measure of the allies' contribution to a fair distribution of the burden (44). In other words, as outlined in the "Report of the U.S. Department of Defense" (45) a few years ago, there is no mathematical formula enabling to combine with appropriate weights all the most important elements of burden sharing in an "optimal" indicator to be used for a division into fair shares of Nato's common defense burden.

6. Conclusions

In conclusion, burden sharing is a problem which is difficult to solve. By relying on data of a certain kind rather than on another - or by expressing them in one currency rather than another - we can prove one thesis and its exact contrary. However, the question of the division of the costs has more than an "objective" aspect: it has a "subjective" aspect which is perhaps even more important.

If the costs and the resources which a country must bear and allocate for national defense and for the collective defense of the Atlantic Alliance are to be such as to guarantee the safety of the country at hand, the latter will depend, as well as on objective data, also on subjective perceptions and evaluations of its defense and attack capability (including that of its allies) and those of the adversary (the force of the Warsaw Pact).

Thus, even if we hypothesize a situation of "loyalty" in the behaviours of all allied countries in reckoning what it is necessary to do for their own security, there can still be differences in evaluation due to several factors: for example, divergences on the gravity of the threat or on how to confront it (more or less diplomacy, more or less military actions) or, yet, on the strategy to use to face this threat (46). On the other hand, it has been recently underlined (47) that the lower share of the expenditure for defense which characterizes certain countries (e.g. Luxembourg, but even Iceland which has no expenses for defense) could express not a behaviour of free riding, but the marginality of the military security of these countries with respect to their military force. For example, Luxembourg's security would not double if this country doubled its military expenses. K. Knorr stressed the importance of a varied set of qualitative factors which, apart from the quantitative ones above discussed, are part

of the problem of burden sharing, and which should be considered when measuring it: these range from the training of the troops, to the army's morale, to the number of available reserves to the measures of efficiency of the armies and of the expenses for defense. Gauging all these factors, however, is extremely difficult and complicated, but this does not make them less important and significant than the other quantitative factors.

In spite of this, the truly decisive element in the question of burden sharing is, in our opinion, that it is impossible to calculate (and discuss) a fair measure in the division of the costs without also considering the differences in the distribution of the "benefits" among the various countries which affiliation to the Alliance involves.

For example, a hefty (or overdue) contribution to the costs of the Alliance could be offset by an equally hefty share of benefits (48). Unfortunately, even calculating the benefits is an extremely difficult task, because these are basically qualitative elements.

Therefore, the fact that an ally bears a larger share than the others allies does not prove, per se, the existence of an "exploitation", since this economic burden could be counterbalanced by a larger share of net benefits deriving from the balance of other types of costs and benefits (49).

In this respect, we can recall the consistent benefits which the United States draw on the economic, political and military level, from its role as leader of the Atlantic Alliance. In the field of military, foreign and economic policy, the positions and the decisions of the United States are the decisive ones most of the time. Also, we could hypothesize the existence of a link between the American contribution and the United States' position as leader of the Alliance. In any case, the data seem to point to a loyal and honest behaviour on the part of the European allies.

K. Knorr (50) underlined instead the possible "price" which the American leadership could inflict on the other countries of the alliance. For example, an overly aggressive or "provocative" foreign and military policy of the U.S. vis-ŕ-vis hostile countries could trigger retaliation, armed conflict or actual wars in which the other Nato countries could be implicated.

Many of these controversial points originate, in our opinion, mainly from the different role which the United States and the European countries have or wish to have in the future on the international level: a global role with global interests the U.S., a much more limited role and profile the European countries. These strategic ambitions are originated by a different perception which the United States and Europe have of the Soviet threat in the world. These differences among the various allied countries seem to confirm the thesis according to which the question of burden sharing, in order to be correctly analysed, calls for a wider approach which goes beyond the question of "who pays how much", but takes into consideration the global set of costs and benefits and, more in general, the divisions of the strategic responsibilities; in other words, the divisions within Nato of the roles and tasks of the various countries.

In conclusion, it seems that the existing studies on the subject does not recognize the Europeans' contributions to the expenses of the Atlantic Alliance in their just measure and in their exact size, while considering their national differences and variability. A number of particularly reliable researches (the "Reports on the allies' contributions to common defense" of the U.S. Defense Secretary Weinberger - e.g. the ones relative to 1982, 1984 and 1985 - and the estimates of Nato's Eurogroup) (51) maintain that "on the whole, the allies seems to bear a fair share of Nato's defense burden".

NOTES

1. Free riding is a behaviour whereby certain individuals or groups of states benefit from the supply of a public good (both domestic or international) while bearing a lower share of the due costs. For a more exhaustive definition of the problem, refer to paragraph n. 4 of this chapter.

2. SAMUELSON P. A., ``The Pure Theory of Public Expenditure'', "Review of Economics and Statistics", Nov. 1954, p. 387.

3. HEAD J. G., ``Public Goods and Public Policy'', "Public Finance", 1962, n. 3.

4. HEAD J. G., "op. cit.", pp. 200-201; HEAD J. G., ``Public Goods and Multilevel Government'', in DAVID W. L. (editor), "Public Finance, Planning and Economic Development: Essays in Honour of Ursula Hicks", McMillan and St' Martin Press, London and New York, 1973, pp. 20-43.

5. HEAD J. G., ``Public Goods and Public Policy'', "op. cit.", pp. 200-201.

6. SAMUELSON P. A., ``Diagrammatic Exposition of a Theory of Public Expenditure'', "Review of Economics and Statistics", Nov. 1955, p. 350.

7. HEAD J. G., "op. cit.", p. 201.

8. HARTLEY K., PEACOCK A., ``Combined Defence and International Economic Cooperation'', "World Economy", June 1978, n. 3, p. 332.

9. ATKINSON A. B., STIGLITZ J. E., "Lectures on Public Economics", McGraw-Hill, Maidenhead, 1980, p. 484.

10. KENNEDY G., "Defense Economics", Duckworth, London, 1983, p. 24.

11. HEAD J. G., "op. cit.", p. 206.

12. ATKINSON A. B., STIGLITZ J. E., "Ibidem".

13. For a brilliant and complete treatment of the problems relative to the existence and supply of international public good refer to FREY B. S., "International Political Economics", Blackwell, Oxford, 1984, (in particular chapter 7). For public goods relative to the monetary field, refer to KRASNER S. (editor), "International Regimes", Cornell University Press, 1983.

14. 0LSON M., ZECKHAUSER R., ``An Economic Theory of Alliances'', "Review of Economics and Statistics", Aug. 1966,p. 267; OLSON M., ZECKHAUSER R., ``Collective Goods, Comparative Advantage, and Alliance Efficiency'', in McKEAN R. N. (editor) "Issues in Defense Economics", National Bureau of Economic Research, N. Y., 1967; SANDLER T., ``Impurity of Defence: An Application to the Economics of Alliances'', "Kyklos", n. 3, 1977, p. 444; KENNEDY G., "Burden Sharing in NATO", Duckworth, London, 1979, p. 25; HARTLEY K., PEACOCK A., ``Combined Defence and International Economic Cooperation'', "World Economy", n. 3, June 1978, p. 332.

15. OLSON M., ZECKHAUSER R., 1966, "op. cit.", p. 268.

16. SANDLER T., 1977, "op. cit."; SANDLER T., FORBES J. F., ``Burden Sharing, Strategy, and the Design of NATO'', "Economic Inquiry", July 1980; CORNES R., SANDLER T., ``Easy Riders, Joint Production, and Public Goods'', "Economic Journal", Sept. 1984; MURDOCH J. C., SANDLER T., ``A Theoretical and Empirical Analysis of NATO'', "Journal of Conflict Resolution", June 1982, MURDOCH J. C., SANDLER T., ``Complementarity, Free Riding, and the Military Expenditure of NATO Allies'', "Journal of Public Economics", Nov. 1984.

17. An "impure" public good is a good that presents various levels of competition and/or excludability. Goods of this type are, for example, information systems, hunting reserves, artificial channels, satellite-transmitted TV programs, etc.

18. SANDLER T., 1977, "op. cit."

I9.SHAFFER S. M., ``Alliance Politics: A Model Based on Divisibility of Payoff'' in LOEHR W., SANDLER T. (editors), "Public Goods and Public Policy", Sage Publications, Beverly Hills, 1978, p. 151.

20. LOEHR W., SANDLER T., ``On the Public Character of Goods'' in LOEHR W., SANDLER T., (editors), 1978, "op. cit.", p. 23.

21. SANDLER T., 1977, "op. cit.", p. 445; KENNEDY G., 1983, "op. cit.", p. 42.

22. SANDLER T., 1977, "ibidem"; LOEHR W., SANDLER T., 1978, "ibidem"; SANDLER T., FORBES J. F., 1980, "op. cit.", pp. 339-341; WHYNES D. K., BOWLES R. A., "The Economic Theory of The State", M. Robertson, Oxford, 1981, p. 178.

23. LOEHR W., SANDLER T., 1978, "op. cit.", p. 27.

24. HYMAN D. N., "Public Finance", Dryden Press, Chicago, 1983, p. 138.

25. JOHANSEN L., ``The Theory of Public Goods: Misplaced Emphasis?'', "Journal of Public Economics", Feb. 1977, p. 147; SCHNEIDER F., POMMEREHNE W. W., ``Free Riding and Collective Action: An Experiment in Public Microeconomics'', "Quarterly Journal of Economics", Nov. 1981, p. 689.

26. BUCHANAN J. M., "The Demand and Supply of Public Goods", Rand Mc Nally, Chicago, 1968, p. 101.

27. BROWNING E. K., BROWNING J. M., "Public Finance and the Price System", McMillan, N. Y., 1979, p. 24.

28. SCHNEIDER F., POMMEREHNE W. W., 1981, "op. cit.", p. 691.

29. LUNN S., "Burden-Sharing in NATO", Routledge Kegan, London, 1983, p. 1.

30. LUNN S., "op. cit.", pp. 2-3.

31. The main objection which the Americans raised was that the European economic growth had not been coupled with an equally intense growth of the resources which the European nations allocated for defense. On this point, see the reconstruction by LUNN S., "op. cit.", pp. 1-3.

32. LUNN S., "ibidem".

33. SULLIVAN L. J., ``A New Approach to Burden Sharing'' "Foreign Policy", Autumn 1985; ROTH W. V., ``After the Nunn-Roth Amendment'', "NATO's Sixteen Nations", July 1985; CONGRESSIONAL DIGEST, "Controversy Over NATO Cost Sharing: Pro e Con", Washington, D.C., Aug-Sept. 1984; GOLDEN J. R., "NATO Burden Sharing", Praeger, N. Y., 1983; RUPP R. W., ``La ripartizione dell'onere della difesa'', "Notizie Nato", Oct. 1982.

34. SULLIVAN L. J., "op. cit.", p. 91.

35. RUPP R. W., "op. cit.", p. 170.

36. LUNN S., "op. cit.", p. 53.

37. DEPARTMENT OF DEFENSE, "Report on Allied Contributions to the Common Defense, A Report to the US Congress", by Caspar W. Weinberger, Secretary of Defense, Washington, D.C., March 1982.

38. RUPP R. W., "op. cit.", p. 172; LUNN S., "op. cit.", pp. 54.

39. RUPP R. W., "ibidem"; LUNN S., "ibidem", pp. 54-56.

40. LUNN S., "ibidem".

41. LUNN S., "ibidem"; RUPP R. W., "ibidem".

42. RUPP R. W., "op. cit.", p. 174; LUNN S., "op. cit.", p. 55.

43. GOLDEN J. R., "op. cit.", p. 26.

44. GOLDEN J. R., "op. cit.", p. 25; RUPP R. W., "op. cit.", p. 174.

45. DEPARTMENT OF DEFENSE, 1982, "op. cit."

46. KNORR K., ``Burden-Sharing in NATO: Aspects of U.S. Policy'', "Orbis", Autumn 1985, pp. 521-522.

47. KNORR K., "op. cit.", p. 518.

48. KNORR K., "op. cit.", p. 520.

49. KNORR K., "ibidem".

50. KNORR K., "op. cit.", p. 519.

51. HUMPHREYS C., ``Telling It As It is'', "NATO Review", Feb. 1986.