NATO MILITARY EXPENDITURE

IN THE POST COLD WAR ERA

Brian Ardy

Centre for European Research, Thames Valley University

Introduction

The end of the Cold War led to the expectation that military expenditure could be reduced. Unlike other conflicts however, the Cold War was in fact an exceptionally prolonged period of peace. So reductions in military expenditure would involve a contraction of expenditure from peacetime levels. This article analyses the changes in military expenditures which have occurred in NATO countries and the implications of these changes. The first section deals with the deficiencies of the data which are considerable in this area. Then the available data on NATO military expenditure is analysed in the second section. The implications of these changes for defence and security in NATO are the subject of the third concluding section.

Data on military expenditure: The problems

All statistics are subject to inaccuracies and limitations, but there are some problems particular to military statistics. There may be reasons to conceal the true extent of expenditure for political or strategic reasons.[1] What is classified as military expenditure may vary between countries[2] or there may be reclassifications over time.[3] Military expenditure is subject to inflation as a result of increases in wages and prices. So money expenditure has to be deflated by some price index but which price index is appropriate?

In this study real military expenditure is obtained by deflating current price figures by the consumer price index.[4] The value of personnel used by the military is measured by their wages. But does this measure the value of the labour to society and does it provide a an accurate measure of the value of military output the labour produces?[5] To compare defence expenditure of different countries requires their conversion into a common currency. The statistics used in this study compare expenditure in US dollars at 1990 prices and exchange rates. With a particularly low value of the dollar at this time this tends to underestimate the overall US share of NATO defence expenditure.[6] Exchanges rates in any case only reflect the value of internationally traded goods. There is a very large part of the domestic economy including the defence sector which is only indirectly affected by exchange rate movements. Where countries enjoy similarly income levels this is not much of a problem but where income levels are different it can make comparisons difficult. The poorer countries in NATO such as Greece and Turkey will have generally lower wage levels than the NATO average. Thus the amount they pay their military forces will similarly be less, thus a comparison of expenditure levels will understate their military efforts, whereas a comparison of troop levels would make for a more favourable comparison.[7]

In view of these problems the analysis of military expenditure statistics must be made with care. Very different conclusions may be drawn when different statistics are used or the same set of data is subject to different statistical adjustments. These difficulties are especially acute when making comparisons between countries. When analysing the data too much weight should not be given to the particular figures thus this article considers broad trends over a number of years.

NATO military expenditure 1985-1995

Defence expenditure yields no direct benefits to the population of the country. Thus the preference to use resources for defence will depend upon the extent of the perceived threat to sovereignty and perhaps the desire for a nation to have the ability to use the ultimate instrument of foreign policy. Two factors have been reducing the propensity to spend on defence: The end of the Cold War and the general desire to reduce public expenditure. Since 1989 the NATO countries have faced a changed security situation and this has had implications for military expenditure. At the November 1991 NATO summit meeting in Rome it was decided to cut NATO military weapons and to reduce the number of NATO forces deployed in Central Europe from 2.8 million in 1991 to 2.1 million in 1994. This meant decreased military expenditure could be justified as being in line with strategic security requirements.

Another contributory factor has been general pressure on Government expenditure which comes from several sources. Firstly the ascendancy of neo-classical economic thought and a belief in the desirability of market rather than government provision. This justified reduced government expenditure and lower levels of taxation to restore incentives. Secondly the desire for reductions in taxation by the electorate associated with a move to the right by a number of parties and with dissatisfaction with the outcomes of government expenditure. Thirdly globalisation undermined tax revenue from companies. Multinationals were able to shift tax liability to low cost locations by for example transfer pricing. In the search for foreign direct investment governments were providing increasingly generous incentives. Finally in an increasingly competitive world governments were reluctant to place tax burdens on companies which undermine their international competitiveness. In these circumstances it was not surprising that military expenditure came under pressure at a time when the level of threat was waning.

The extent of a nation's defence expenditure depends upon two factors: Firstly, the size of its economy which can be measured by Gross Domestic Product (GDP). Secondly, the strength of the government's preference for defence which is indicated by the percentage of GDP spent on defence. These two factors mean that the USA dominates NATO defence expenditure accounting for 59.0% of the 1994 total.[8] This was because US GDP amounted to 46.8 per cent of NATO GDP (World Bank, 1996, p.211) and it spent an above average proportion of its GDP on defence in 1994, 4.3% compared with the NATO average 2.4%. The major European countries spend a smaller proportion of their GDP on defence: France 3.3%, UK 3.3%, Italy 2.0% and Germany 1.8% in 1994. The size of the German, French, Italian and UK economies mean that they dominate European NATO defence expenditure accounting in 1994 for 78.4% of the total. Thus the US, France, the UK, Germany and Italy accounted in 1994 for 89.3% of total NATO defence expenditure. So it is what happens in these countries especially the USA which determines overall trends in NATO military expenditure.

Table 1.

Real NATO military expenditure 1985-1995 (Index 1987=100)

Sources: SIPRI (1995), Table 12A.2, p.440; SIPRI (1996) Table 8A.2, p.365; NATO Review (Various).

Real NATO military expenditure peaked in 1987 (See Table 1) but it was only after 1989 that real expenditure began to be reduced substantially below its 1987 level. The 1987 peak is the result of spending trends in the USA. The first half of the 1980s showed a substantial increase in real expenditure associated with the Reagan Presidency. Real US defence expenditure increased 39.8% from 1980-86, with in 1986 6.6% of GDP spent on defence. Associated as it was with a very large federal deficit this level of defence expenditure was unsustainable. So reductions in US defence expenditure initiated the downward trend in alliance defence expenditure. After 1989 these cuts in defence expenditure could be further justified by the reduction in threat and the implementation of arms reduction agreements.

Perhaps not surprisingly expenditure in the UK follows a similar pattern to the USA but with an even earlier peak in 1984. The UK reductions in expenditure have been driven by a desire to reduce public expenditure, with defence an obvious target as a result of its previous expansion. The UK at least partly as a result of this situation, was quick to seek to adapt its defence posture to the realities of the post Cold War era with 'Options for Change' announced in 1990 (King, 1990). Although like other NATO countries the UK still finds difficulty in defining its new military role it has at least been under consideration and as a consequence expenditure reductions have been undertaken in a reasonably well planned manner. (Taylor, 1994). It was not until 1992 that European defence expenditure was cut significantly below its 1987 level. US and NATO expenditure rose in 1992 is a result of the Gulf War. By 1995 real NATO expenditure is estimated to have fallen to 77.0% of its 1987 peak. Despite the fact that defence spending has been cut faster in the USA than in Europe in real terms US military expenditure was still higher in 1994 than it was in 1980. By contrast in NATO Europe real expenditure in 1995 was below that of 1980.

The reduction in expenditure represented an even more significant reduction in 'defence effort' as measured by the proportion of GDP spent on defence. For NATO this declined continuously over this period from 5.2% in 1985 to 3.2% in 1994 a reduction of over a third. The USA reduction at 36.4% was even greater from a 1986 high of 6.6% to 4.2%. European countries devote much less of their output to defence but there was still a substantial reduction in effort from 3.5% in 1985 to 2.4% in 1994 (-31.4%).

Table 2.

Defence expenditure as % of GDP 1985-1995 (% of GDP at current prices)

* 1995 estimated

Source: NATO Review (Various)

It is possible to place NATO members, roughly, into three categories according to the recent trend in their real defence expenditure: Expenditure reducing, expenditure maintaining and expenditure increasing. In the first and largest category significantly reducing expenditure are Belgium,[9] USA, UK, Germany, Spain, Canada and the Netherlands. The second category, with real expenditure near to 1987 levels, includes France, Italy, Denmark, Greece and Norway. The only three countries which have significantly increased defence expenditure in real terms since 1987 are Portugal, Luxembourg and Turkey.

These different patterns of expenditure reflect a number of factors. The USA and the UK increased spending rapidly in the early 1980s (spending has since declined from this peak). Unlike the US, however, the expenditure of the UK is well below that of 1980. This is also the case for the other European 'defence expenditure reducers' including Germany. The reduction in German defence effort is even more significant than the figures suggests because with unification in 1990 Germany acquired significant additional forces and defence responsibilities. Thus German defence expenditure rose in 1990 but by 1991 it had been cut below its 1989, pre-unification, level. These German reductions are justified by the lower level of threat to a former NATO front-line state but are also the result of budgetary considerations. Germany is being affected by a particularly severe budgetary squeeze caused by the recession, unification, very large external commitments and by the need to meet the Maastricht convergence criteria for Monetary Union. Further cuts in German military expenditure below planned levels were made in 1996 and more may well be required in 1997 (Lindemannn, 1996). Except for unification, the same factors are also influencing defence expenditure in other EU NATO members.

Italian expenditure was highest in 1989, with significant reductions in 1990 but further significant reductions were only achieved in 1995. Italy's defence policy was affected by the changing international environment and by the need for budgetary stringency. Reductions in expenditure were limited, however, by the slow progress of the Italian government in drawing up its New Defence Model. Additional factors were the comparatively low level of defence expenditure, the high proportion of expenditure on manpower and an expansion of the Carabinieri (Silvestri, 1994).

By contrast French expenditure has been relatively stable. To a significant extent this seems to reflect procrastination, especially in the last few years of the Mitterrand Presidency and 'cohabitation' between the socialist president and Gaullist prime minister (Paolini, 1994). Thus the new Government is proposing substantial cuts in expenditure (Buchan, 1996). Among the other NATO members Greece and Turkey really belong to a different geostrategic theatre, especially since their expenditure levels are related, to a large extent, to tensions between them (Chletos & Kollias, 1995; Kollias, 1996). A further indication of their higher perceived level of threat is that their defence effort measured by the percentage of GDP devoted to defence is the highest in Western Europe Turkey 4.2% and Greece 5.6% in 1994.

NATO subdivides military expenditure into four areas: Equipment, personnel, infrastructure and other operating expenditure.[10] The difference between these categories is not entirely clear (SIPRI, 1994, p.402). Equipment is presumably largely made up of expenditure on weapons systems and other types of (non-building/ base) capital. Personnel would include expenditure on wages, salaries and pensions charged to the military budget. Infrastructure would be spent on the purchase of land, the construction of buildings, runways etc. Not surprisingly infrastructure expenditure is generally low, less than 5% of the total, but it does fluctuate considerably. By a process of elimination other operating expenditure would be operations and maintenance spending. Care has to be taken in the interpretation of the data because to some degree the classification may be arbitrary. There are particular problems with the rather ill-defined 'other operating expenditure' –for example, the US Operations and Maintenance budget in 1994 included 'aid to the former Soviet Union' and other 'Global Cooperation Initiatives' (SIPRI, 1994, p.404). In addition changes in the way in which tasks are carried out may affect classification –for example, contracting out services to private companies might shift expenditures from personnel to other operating expenditure. This subdivision of the budget does reveal some particular differences between countries.

Table 3.

Real Defence Expenditure NATO Countries (Real expenditure in the national currency as a % of 1987)

Source: SIPRI (1996) Table 8A.2., p.356.

For all countries personnel accounts for the largest proportion of expenditure. However, the share varies enormously between NATO members (from 69.7% in Belgium to 34.9% in Norway in 1995). This may be the result of differing classification practices with the countries with low personnel expenditure, such as the US and the UK, having higher 'other operating expenditure'. Lower personnel expenditure may also reflect the greater priority given to equipment in some NATO members. Personnel expenditure will depend upon the size of military forces, the level of pay and other liabilities such as pensions. Although France, the UK, Germany and Italy accounted for over three quarters of NATO's European defence expenditure they only provided around half of Europe's military personnel. The largest armed forces in Europe are those of Turkey which, with over 800,000 military personnel, are now just over half the size of those of the US. Greece has the sixth largest number of military personnel in Europe. The size of the Greek and Turkish armed forces is another manifestation of their different geostrategic situation and of the tensions between them. It is intersting to note that the proportions of their labour force employed in the military are by far the largest in NATO, with the 1994 figure for Greece at 5.6% and that for Turkey at 4.2%, compared with a NATO average of 2.0%.

Reductions in troop levels have not been matched by reductions in personnel expenditure. There is a tendency for the wages of the armed forces to rise faster than prices in line with increasing wages in the general economy. The growing professionalisation of the armed forces has increased this tendency. In addition reductions in force levels lead to extra costs some of which are met from the personnel budget. The biggest discrepancy is for Germany where troop levels fell by 26.9% between 1987 and 1994 but personnel expenditure (in DMs) was only reduced by only 3.8%. This reflects rising real wages in Germany. In Turkey, Portugal and Luxembourg real personnel expenditure 1987-94 has increased substantially, more than doubling in Turkey.[11] Personnel expenditure has remained at around its 1987 level in a number of countries, namely Italy, Denmark, Spain, Greece and Germany. The only countries to have significantly reduced their personnel expenditure are Belgium, Norway, UK and the USA.

Table 4.

NATO Armed Forces

Numbers in thousants - % change from 1987

* From 1993 conscripts are excluded from Norwegian force numbers.

** From 1992 Gendarmerie are excluded from Belgium's force numbers.

Source: NATO (Various) Defence Expenditures of NATO Countries, NATO Review.

Capital expenditure on equipment is relatively easy to reduce, simply by failing to place new contracts. The most important countries for equipment expenditure are the USA and UK.[12] The US makes a particular contribution to the alliance with its military hardware, such as nuclear weapons. Agreements on strategic arms limitations have had a particular impact on US military expenditure, enabling the US to reduce equipment and other operating expenditures. In the US and the UK expenditure on equipment had expanded rapidly in the 1980s. With forces comparatively well provided with new equipment it was possible for a while to reduce substantially this expenditure. Thus, the recovery of equipment expenditure in the US and the UK in 1994 and 1995, may reflect the end of the period when it was possible 'to live off the fat' of previous increases in defence expenditure. By 1993 equipment expenditure had been reduced to 67.5% of its 1987 level in the US but in 1994 it had risen back to 84.5%. In the UK equipment expenditure in 1992 was only 63.9% of its 1987 level but by 1993 it had risen to 89.8% . Generally, equipment expenditure was cut particularly deeply between 1987 and 1995 with most NATO members reducing their expenditure by more than a quarter: Belgium (-64.7%), Spain (-54.1%), Germany (-57.6%), Netherlands (29.9%), Italy (-28.2%) and Canada (-25.6%), USA (-24.8%) and the UK (-12.3%).

Unlike the USA and the UK cuts are projected to continue in a number of countries. Although in ministerial statements Germany continues to affirm its commitment to the Eurofighter project there is no allocation in the 1997 budget for the project (Fisher, 1996). Danish equipment expenditure was maintained and the only countries to have increased their real expenditure on equipment, in 1987-95, are Turkey (+120.9%), Portugal (+78.5%), Luxembourg (+66.6%),[13] Norway (+21.7%) and Greece (+14.5%).

Other operating expenditures have also been severely squeezed. The cuts, in 1987-1995, do vary between countries. For example, the UK (-47.9%), Turkey (&endash45.8%), Spain (-41.5%) USA (-37.4%), Belgium (-30.2%), Germany (-30.0%) and Canada (-25.6%) all reduced their other operating expenditures by more than a quarter. Cuts were less severe in Greece (-22.8%), Norway (-21.5%), the Netherlands (-21.5%) Denmark (-6.5%) and Italy (-9.5%). Portugal (2.5%) maintained its expenditure. These cuts reflect the lower costs of supporting fewer troops and less hardware. In the case of the US the reduction in its strategic arsenal had a particular impact on other operating expenditure. With reductions in troop and equipment levels, cuts in operations and maintenance are to be expected. So it is difficult to predict the effects of these cuts on the preparedness of the remaining forces.

Conclusion

NATO members' military expenditure has fallen substantially in real terms and even more as a proportion of GDP. That the reductions represent a real fall in effort is indicated by the reductions in force levels and by the pressure investment programmes are under. Further reductions, although justified by the lowered level of threat, are often part of more general government expenditure reduction programmes, so cuts in military expenditure seem likely to continue. These reductions in defence expenditure seem to have preceded the development of plans for defence requirements in the post-Cold War World. Although the strategic situation is by no means resolved three characteristics have emerged. Firstly, the development of professional force and the end of conscription. Secondly, greater co-operation among military forces particularly in Europe. Thirdly, far greater commonality of procurement for equipment and for a restructuring of the defence industry.

There are some important advantages in conscription. Some countries see conscription as way to achieve a citizens' army not divorced from the population as whole. A conscript army may be cheaper because to ensure a sufficient number of recruits to an all volunteer army would require a significant increase in the wages of the armed forces. Thus a Dutch Ministry of Defence report considering conscription suggested that an all-volunteer force would be too expensive and would find it difficult to ensure sufficient recruits (Doel, 1994, p.63). The UK forces despite their experience still have difficulties attracting and retaining an adequate supply of recruits. However, despite these advantages, the viability of conscription is rapidly diminishing. The possibility of a large conventional war in Europe has become remote. There is an increasing need to use troops out of NATO areas, particularly in peacekeeping because of increasing instability. A conscript army might be suitable for the defence of their own country but sending them abroad on hazardous missions peripheral to the national interest is controversial. The formation of all-volunteer units for overseas missions would be logistically difficult. A further problem is that as force levels are reduced, either the length of the period of conscription is so shortened as to become meaningless or conscription becomes rather a lottery. Thus in the Netherlands only 38% of approved conscripts are actually serving (Doel, 1994, p.63). Military service is also increasingly unpopular, in Germany, in 1995, 160,590 youngsters opted for civilian service instead of military service (Lindemann, 1996). These pressures have already led the Netherlands to abandon conscription and France to plan to phase it out by 2002. Conscription is thus likely to be ended in most NATO countries relatively soon.

Table 5.

Real NATO military expenditure 1985-1995 (billion US$ 1990 prices and exchange rates – SIPRI, 1996)

Source: SIPRI (1995) Table 12A.2, p.440., SIPRI (1996) Table 8A.2, p.365, NATO Review (Various).

Shrinking armed forces are becoming increasingly stretched in trying to retain their effectiveness across a spectrum of operational requirements. With the very high costs of modern equipment meeting declining defence budgets, the number of ships, planes and tanks that forces can buy and operate is diminishing. Thus it is becoming more and more difficult for countries to maintain the competence of their forces. The need to operate overseas multiplies the difficulties as the costs of such operations are high and the transport and logistics required are formidable. In the longer term the only way to maintain effective forces with shrinking budgets is for countries to pool their forces. There is already substantial co-operation but logic dictates that this needs to go further with combined forces. With continuing reductions in defence expenditure European military forces will only be able to offer effective security when they are operated jointly. These pressures together with fears of US disengagement from Europe have led to France's partial reintegration into NATO's military structure and the attempts to develop a European military identity within NATO based on the WEU (Cornish, 1996). The WEU has significant gaps and deficiencies in its defence capabilities in:

• crisis management mechanisms;
• reconnaissance and intelligence;
• transport capabilities;
• standardisation and interoperability; and
• the European defence industrial base (WEU, 1995).

The substantial reduction in defence expenditure means that these problems can only be effectively addressed within a much higher level of integration than is being achieved at the moment.

At one time a comprehensive indigenous defence industry was seen as an essential component of security, a vital part of the national industrial base and an important source of export earnings. With escalating R&D costs for the increasingly sophisticated weaponry, European defence companies were co-operating more and more even during the Cold War. The defence industry, in NATO countries, has been affected not only by reductions of NATO demands but also by shrinking trade in arms beyond NATO with the US, the UK, France and Germany all being major arms exporters. Non-Soviet arms exports fell by 27.5% between 1987 and 1994[14] (IISS 1996). These developments have two implications. Firstly, that the European defence industry will have to become more concentrated; and secondly, that procurement of weapons systems will have to be joint among a number of countries. Collaborative projects have of course been common for a considerable period. Formal links are now being developed with the establishment of an agency to oversee joint arms production projects by France and Germany, which is now to be joined by the UK and Italy (Clark, 1996). Reductions in defence equipment expenditure and shrinking export markets have forced even the already large US weapons companies to concentrate their activities with 'mega-mergers' between Northrop and Grumman, and Lockheed and Martin (SIPRI, 1996, p.419). European companies are only just beginning the process of consolidation and there are considerable difficulties in the way of this process. The regrouping has been on a largely national basis with two defence contractors in the UK, one in Germany and one in Italy. In France the process is just beginning with seven big contractors still existing but the preference for national solutions persists, e.g. the merger between Aersopatiale and Dassualt. The difficulties of cross-border rationalisation is illustrated by the development of the British Aerospace/Matra missile joint venture (Gray, 1996). The retention of Article 223(b) of the Treaty of Rome means that EU governments can protect their arms manufacturers avoiding the normal rules which govern state aids, procurement and mergers. The ultimate beneficiaries of this tardiness in restructuring the European defence industry may well be the US manufacturers (Economist, 1996b).

In the revision of strategy that followed the end of the cold war NATO agreed to reduce substantially nuclear weapons, to reduce troops committed to Central Europe and to reorganise the remaining forces. These changes recognised that the security in Europe no longer required a very large conscript army –backed by nuclear weapons– prepared to fight a war against the Warsaw Pact on home soil. Military forces were still required to meet a residual threat from the East, as well as the growing demand for peacekeeping for civil wars and localised conflicts in Europe and beyond. Additionally there was the possibility of other 'Gulf' type conflicts in which some countries, particularly the UK and France, might wish to participate.

As always with military expenditure it is difficult to relate these new requirements to force and expenditure levels. Whilst these changed demands undoubtedly permitted a reduction in force levels they also required significant changes in the organisation, training, equipment etc. Some of these changes, such as the reduction of military personnel and lower demand for nuclear weapons, led to reductions in expenditure. Others, such as the increased need for volunteer troops and their greater mobility, would be likely to increase expenditure. Whilst overall expenditure could be reduced it is unclear by how much, if the required capabilities are to be retained. The reductions in military expenditure documented in this article are such that the extent to which NATO countries, particularly in Europe, retain the capabilities which NATO's revised strategy requires is now open to question. In the absence of a credible overall threat to the Alliance and the resulting diminution of the US defence commitment to Europe, reductions in military expenditure and their uneven distribution could well undermine the cohesiveness of the alliance.

_______________________________________________

ENDNOTES

[1]. The £1 billion spent on the development of a new warhead for Polaris missiles from 1975-91 did not appear in military expenditure. (Blackaby & Ohlson, 1987, p.13) This expenditure was concealed because of splits in the Labour Party over expenditure on nuclear weapons.

[2]. A number of countries include effectively civilian police under military expenditure, e.g. the Carabieneri in Italy (Silvestri, 1994, p.52).

[3]. The cause of a significant drop in military expenditure in Belgium in 1992 was the result of the decision to demilitarise the Gendarmerie (NATO, 1996, p.32).

[4]. If an assessment is being made of the effort put into military activity the consumer price index would be suitable. Such an index would measure the real value of civilian activities given up for defence expenditure over time. If the purpose of the analysis is to measure changes in defence output then using the consumer price index assumes that the rate of productivity increase in military activity is the same as that in civilian industry. This might not be appropriate since defence equipment is very research and development (R&D) intensive. R&D can account for 30-40% of the cost of military equipment compared with 1-3% of the value of manufacturing output (Blackaby & Ohlson, 1987, p.19). This R&D if translated into greater effectiveness would mean that output was increasing much more than expenditure (the productivity of capital equipment is rising). Given the difficulties of measuring military productivity the correct extent of the adjustment is difficult to gauge, but given the extent of the R&D effort the price correction for military equipment expenditure should be lower than that for consumer expenditure. 'There is thus no single unambiguous figure for measuring the volume increase of military expenditure: it depends on the price index used.' (Blackaby & Ohlson, 1987, p.21).

[5]. With conscript forces defence effort may be underestimated if the wages of conscripts are below market levels. What however, is the productivity of conscripts compared with volunteer soldiers? Professional forces remain in military service for much longer periods than conscripts the training costs for an equivalent force are lower. With a conscript army the training costs most of which are labour costs are hidden in the general personnel budget. Professional forces should also be better trained as they can over time acquire a range of skills. Another improvement in productivity relates to the greater flexibility in terms of location and type of operations upon which professional forces can be sent. So some at least of the higher costs of professional forces will be offset by higher productivity.

[6]. Against the ECU the US dollar depreciated by over one third between 1985 and 1990 and 1990-94 it appreciated 7%. So using 1990 exchange rates reduces the relative size of US defence expenditure substantially compared with 1985-1986 exchange, more marginally compared with 1987-89 or 1993-94 exchange rates. Changes in real defence expenditure for a country will be unaffected by the currency in which the expenditure is expressed provided only one exchange rate is used for the whole period.

[7]. Whether relative troop levels overstate the contribution of poorer countries depends upon the productivity (quality) of their forces compared to NATO forces in richer countries. Poorer countries' troops are likely to be less effectively armed. This is because the relative abundance of labour and scarcity of capital in poorer countries means that a poorer country will tend towards personnel intensive methods of defence, whereas a richer country will use more capital (equipment) intensive methods (Fontanel, 1987, p.30).

[8]. All the statistics in the text of the article are taken from SIPRI (1996) except where otherwise specified.

[9]. Belgium's reduction in expenditure is overstated because in 1992 the Belgium Government demilitarised the Gendarmerie.

[10]. Figures are not provided for France because of it is not part of the integrated military structure. Since France has the second largest military expenditure in NATO this means that overall NATO expenditure cannot accurately be divided into the different categories. Thus only tentative conclusions can be made about trends in these categories of expenditure across the alliance.

[11]. Care has be taken in interpreting Turkish expenditure levels because of the high and unstable rate of inflation and the instability of the exchange rate. Thus Turkish personnel expenditure is estimated to have been $3169m. in 1994 but $2295 in 1995.

[12]. Excluding France for which no breakdown of expenditure is available. France is likely to be the second largest equipment purchaser.

[13]. Portuguese and Luxembourg equipment expenditure is relatively small and very unstable.

[14]. Including Soviet exports the trade in arms fell by 62.3% from $78.6bn in 1987 to $29.6bn in 1994. Soviet exports of arms however, fell from $29.1bn in 1987 to $1.3bn in 1994. There was a marginal recovery of arms exports to $30.2bn in 1995.

BIBLIOGRAPHY

Blackaby, F. & Ohlson, J. F. (1987) 'Military Expenditure and the Arms Trade: Problems of Data', in Schmidt (ed.) (1987).

Buchan, D. (1996) 'French cabinet approves defence cuts', Financial Times, 14 May.

Chletos, M. & Kollias, C. (1995) 'The demand for Turkish military expenditure 1960-92', Cyprus Journal of Economics, Vol.8, No.1, Vol.8, pp.64-74.

Clark, B. (1996) 'UK and Italy join European arms agency', Financial Times, 12 November, 1996.

Cornish, P. (1996) 'European Security: the end of architecture and the new NATO', International Affairs, Vol.72, No.4, October, pp.751-769.

Doel van den, M. (1994) 'Dutch Defence Policy', in Taylor (ed.) (1994).

Economist (1996a) 'NATO acquires a European identity', The Economist, 8/6/96.

Economist (1996b) 'Europe's defence industry', The Economist, 17/2/96.

Fisher, A. (1996) 'Waigel reaffirms Eurofighter plans', Financial Times, 11/11/96.

Fontanel, J. (1987) 'A Note on the International Comparisons of Military Expenditure', in Schmidt (ed.) (1987).

Gray, B. (1996) 'An elusive moving target, Financial Times, 14/5/96.

International Institute for Strategic Studies (1996) The Military Balance, IISS, London.

Kaufmann, W. W. (1994) ''Hollow' Forces', The Brookings Review, Fall, pp. 24-29.

King, T. (1990) 'Statement by Tom King Secretary of State for Defence, to House of Commons', 25 July 1990, Hansard, vol. 177, no. 150, Col. 470.

Kollias, C. G. (1996) 'The Greek-Turkish conflict and Greek military expenditure 1960-92', Journal of Peace Research, Vol.33, No.2, pp. 217-228.

Lindemann, M. (1996) 'Defence to bear brunt of Bonn cuts', Financial Times, 29/5/96.

NATO (1994) 'Defence Expenditures of NATO Countries' (1970-1993), NATO Review, April, pp. 33-35.

NATO (1996) 'Defence Expenditures of NATO Countries' (1975-1995), NATO Review, January, pp. 31-33.

Paolini, J. (1994) 'The French Case', in Taylor (ed.) (1994).

Schmidt, C. (ed.) (1987) The Economics of military expenditures: military expenditures, economic growth and fluctuations, Macmillan, Basingstoke.

Silvestri, S. (1994) 'Italian Defence Policy', in Taylor (ed.) (1994).

SIPRI (1994) SIPRI Yearbook 1994, Stockholm International Peace Research Institute, Oxford University Press.

SIPRI (1995) SIPRI Yearbook 1995, Oxford University Press.

SIPRI (1996) SIPRI Yearbook 1996: Armaments, Disarmament and International Security, Oxford University Press.

Taylor, T. (1994) 'British Defence Policy', in Taylor (ed.) (1994).

Taylor, T. (ed.) (1994) Reshaping European Defence, Royal Institute for International Affairs, London.

WEU (1995) WEU Council of Ministers, European Security: A Common Concept of the 27 WEU Countries, Madrid, 14 November. Extracts in SIPRI (1996) pp. 309-312.

World Bank (1996) From Plan to Market, World Development Report 1996, Oxford University Press.