The ideas which are here expressed so laboriously are extremely simple and should be obvious. The difficulty lies not in the new ideas, but in escaping from the old ones, which ramify … into every corner of our minds.
J M Keynes

South Africa’s military expenditure

Between 1990 and 1999, South Africa’s military expenditure (milex) fell by some 57% in constant prices. In 1999, it made up around five per cent of all government expenditures. The vast majority of this milex goes on current expenditure, particularly personnel costs, rather than on capital equipment or arms. Personnel figures are being reduced to around 70,000.

A perspective on these figures can be gained by comparing relevant indicators among the Southern African Development Community (SADC) member countries and these are summarised in Table 1 (over page). In 1999, South Africa spent a similar amount on the military as the rest of SADC combined. For 1997, the most recent year for which comparable data are available, South Africa had a smaller number of military personnel than the rest of the SADC countries combined but a vastly superior force projection, as indicated by the respective numbers of ‘heavy weapons’.

This decline in milex was reversed by the decision in late 1999 to spend some R30 billion (around US$4.8 billion) on a number of ships and aircraft. More recently, the army has revealed its intention to acquire tanks, other vehicles and ground to air missile systems worth R16 billion. These acquisitions involve extremely sophisticated weapons and will mean that the military superiority of South Africa in the region will increase dramatically. Inflation and exchange rate adjustments have meant an increase in the cost of the arms package, according to the 2002 Budget Review, to R52.7 billion. Other estimates indicate a much higher figure.

Table 1: Military indicators for Southern Africa

Much of the cost will be met by government borrowing from foreign commercial banks, but there will also be an increase in the national budget allocation to defence, to over six per cent in the 2001 and subsequent budgets.

This article will evaluate South Africa’s military expenditure in four ways:

by reference to the country’s security needs;


by comparison with national defence policy and principles;


by examining its impact on economic growth and development; and


by considering its cost effectiveness in achieving security.

South Africa’s security needs

The nature of armed conflict and the meaning of security have changed significantly in the past two decades. First, almost all armed conflicts now occur within countries rather than between them, typically between government forces and groups wishing to secede or take over government. In 1999, for example, there were 27 ‘major armed conflicts’, defined as:

prolonged use of armed force ... incurring the battle-related deaths of at least 1,000 people during the entire conflict and in which the compatibility concerns government and/or territory.2

Only two of these 27 were interstate wars, of which one—between Ethiopia and Eritrea—was in Africa. Worldwide, invasions of one country by another are very rare events. Of the remaining 25 internal conflicts, ten were in Africa (Algeria, Angola, Burundi, the Republic of Congo, the Democratic Republic of Congo—DRC, Guinea-Bissau, Rwanda, Sierra Leone, Somalia and Sudan). None of these are in countries which have borders with South Africa and there was an average of 38,000 battle deaths from wars in sub-Saharan Africa in the three years 1997 to 1999, 45,000 of which occurred in Ethiopia-Eritrea. To provide some perspective on these figures, it may be noted that around 250,000 people are dying each year of AIDS-related illness in South Africa and a further 60,000 from road accidents and interpersonal violence.

Related to the changed nature of wars, the traditional definition of security has, over the past 20 years, come to have limited relevance to developing countries. The traditional definition focused on territorial security—on external military threats to nation states— and military power to deter or resist such attacks was seen as the prerequisite of national security. Those arguing for a new concept of security suggest that the type and source of threats now facing developing countries are much wider and more complex. For example, the UN Development Programme3 has argued for a redefinition, which focuses on human security, under which it identifies seven aspects—economic, food, health, environmental, personal, community and political security. It points out that a person in a developing country is 33 times more likely to die as a result of ‘social neglect’ (preventable disease and malnutrition) than as a result of an inter-country war.4

Our way of thinking about security, then, needs to change from an emphasis on territorial security based on a strong military to human security, based on improving levels of human development. Logically, if issues of human security and human development are of far greater importance than territorial security, the expenditure of R16.1 billion on defence in the 2002 budget denies resources to other, more important needs. This point is dealt with again later.

Another approach to redefining security comes from peace theory, which extended the concept of violence from direct or physical violence to structural violence.

Johan Galtung coined the term structural violence as a result of fieldwork in colonial Rhodesia where he became increasingly aware of the limitations of defining peace as simply the absence of violence. He noted that while there was little direct violence by the colonial authorities against the native population, there were structures that had significant negative effects on African people:

In a certain sense, there was harmony, co-operation, integration. But was this peace? With the blatant exploitation, with blacks being denied most opportunities for development given to whites, with flagrant inequality whereby whites were making about 20 times as much for exactly the same job as blacks? Not to mention the basic fact that this was still a white colony.5

Structural violence, then, describes those systems that maintain the dominance of one central group over another group, often the majority, at the periphery. Quietly and unintentionally, structural violence nonetheless kills 30 to 40 times more people than does war.6 ‘Negative peace’ involves cessation of physical violence while ‘positive peace’ occurs when the underlying causes of the conflict—exploitation, inequality and the like—are dealt with.

In the context of a changed security environment, and in the understanding of what security means, it seems that the size and composition of South Africa’s military expenditure is irrational. There is a need to move to new ways of thinking about what security means and for this thinking to be reflected in decisions about military expenditure.

Security needs and defence policy

Has the South African National Defence Force (SANDF) adapted to these new ways of thinking or is it locked into the old ways? At a policy document level, the Department of Defence appears to accept this new thinking, as indicated by the following quotation from its website:

The greatest threats to the people of South Africa at present are socio-economic problems such as poverty, unemployment, lack of education, lack of housing, the high crime rate and violence. One of the policy priorities of the Government, therefore, is socio-economic development and upliftment.7

However, despite this statement, the present and proposed future size and composition of the SANDF are oriented towards armed conflicts of a kind that are extremely unlikely to occur. South Africa is still locked into an out-of-date, territorial concept of security that, as we shall see, imposes heavy costs on the economy and society.

Further evidence of variance between policy statements and practice is evident from a reading of the White Paper on National Defence for the Republic of South Africa of 1996 and the subsequent Defence Review of 1997. Consider the following quotations:

“The government has adopted a broad, holistic approach to security, recognising the various non-military dimensions of security and the distinction between the security of the state and the security of people. The greatest threats to the security of the South African people are socio-economic problems like poverty and unemployment, and the high level of crime and violence …”



“The government has adopted a narrow, conventional approach to defence. The primary function of the SANDF is defence against external aggression. The other functions are secondary …”



“South Africa is committed to the goals of arms control and disarmament at national, regional and international levels …”

Taken as a whole, and in the absence of any conventional threat, these statements suggest at most a very modest defence force with a strongly defensive orientation. Logically, the appropriate size of the ‘core force’ in terms of milex, personnel and heavy weapons would be arrived at by reference to the military strength of the countries most likely to engage in aggression against us. As we have seen, South Africa already totally dominates the region militarily, even before the addition of the planned arms purchases. No country in the region poses the remotest threat to South Africa. There is therefore a serious lack of compatibility between national defence policy and subsequent practice.

If there are no current threats, what about the future? In terms of the traditional view of national security, the SANDF admits that there is no foreseeable military threat to the country but goes on to argue that:

… the essential unpredictability of international relations, in which an unforeseen threat might materialise relatively quickly, together with internal potential for disorder, means that we need to retain a capability of defence against lawlessness—whether internal or external … Although our peacetime defence force does not need to be as large as a wartime force, it cannot afford to become complacent. A technologically advanced, well-trained defence force is the pride of any country and a powerful deterrent to would-be aggressors in the troubled world we live in. It must maintain its preparedness, morale, equipment and its capabilities at a high level so as to be ready for any possible threat to the territorial integrity of the Republic.8

And from the Defence Review:

Deterrence requires the existence of a credible defence capability to inhibit potential aggressors. Although South Africa is not confronted by any foreseeable external military threat, this capability cannot be turned on and off like a tap. It is therefore essential to maintain a core defence capability with the potential for expansion should this ever be necessary.

Such justifications of the size and sophistication of the SANDF are based on extremely unlikely future possibilities. Of course, it is always possible that the government of some currently passive country may become aggressive and attempt an invasion. The questions which need to be asked are how likely is this possibility and whether it would occur unexpectedly or only after some observable preparations. There is no country that has both a motive and capability of posing a military threat to South Africa. In this respect, South African defence policymakers seem to be emulating their US counterparts who contend that the US needs the military capacity to be able to fight two major wars at one time. In fact, it has been involved in only three major wars in the 55 years since the end of the Second World War and these occurred decades apart.9

Perhaps the weakness of the case for heavy milex is apparent in defence circles. It is interesting to note the suggestion made at the 2000 African Defence Summit that the primary role of the SANDF be changed from defence against external threats to the support of peace operations and diplomatic peace initiatives and co-operation with other states.10

First, if such a change is to be made, it is imperative to reconsider the composition of the arms package, very little of which would be useable in peacekeeping operations. Second, it is important not to over-estimate our capacity to contribute to such operations. It has been estimated that as many as 80,000 peacekeepers would be needed for the DRC, whereas South Africa could provide a maximum of perhaps 2,000 properly trained personnel. The SANDF deployment of 700 peacekeepers to Burundi and 120 to the DRC indicates the country’s intentions as regards peacekeeping in the region.

Thus far, this paper has attempted to demonstrate that South Africa’s military expenditure is irrational in terms of the security situation the country is likely to face over the next 20 years or so and is not compatible with significant parts of stated security policy. We now turn to an area of specific interest to economists—the relationship between milex and economic growth.

The benefits and costs of the military

The broad benefits and costs of the military (and therefore milex) have been summarised in Table 3. All of these benefits and costs have applied to some country or another at some time in their histories. The fundamental benefit, which justifies the maintenance of a military force, is security from external attack, the validity of which for South Africa has already been discussed. The others, according to the Defence Review, are of secondary importance. The key question for the economist, on which we focus, is the net effect of milex on economic growth. Before doing so, benefits 5 to 9 inclusive will be commented on.

Table 2: Major arms purchase components, 2000–2006

Table 3: Benefits and costs of military expenditure (milex)

There is no doubt that a strong military builds national pride, creates jobs, etc. The economist must examine how cost effectively it does this. In the case of jobs, for example, it has been asserted that the new arms purchase arrangements, which involve offset agreements by foreign companies to invest in South Africa, will create 65,000 new jobs. The economist’s question is ‘fine, but at what cost?’ From a job creation perspective, the spending of R52.7 billion to create 65,000 jobs means that each job costs more than R800,000, which could employ a dozen teachers or nurses. Milex is invariably an extremely cost-ineffective way of creating jobs.

His early research on the relationship between military expenditure and economic growth (cost 4) in developing countries led Emile Benoit11 to conclude that military expenditure promoted economic growth, principally by enhancing human capital. Benoit was wrong. Subsequent research, of increasing econometric sophistication, has almost always reached the opposite conclusion. In national accounting terms, military expenditure is a consumption item, and does not provide a stream of returns in the future. As such, it reduces government saving and thereby reduces the resources available for government investment; economic growth is thus retarded. This negative effect is generally so large as to completely overwhelm any positive effects.12

A related question is whether high military expenditure encourages private investment—both domestic and foreign—or retards it. Presumably, if investors perceive a strong military as having a stabilising or growth-enhancing influence on the economy, thus making high returns more likely, high military expenditure will encourage investment. Alternatively, they may view the military as a destabilising influence and invest elsewhere. The evidence for South Africa is that milex has no significant effect on private investment; other determinants are much more important.

A second way in which military expenditure may negatively affect economic growth is via government expenditure tradeoffs. While studies have often been inconclusive, it is obvious that military expenditure has opportunity costs and may constrain other, possibly more productive, government expenditures. Examples of significant initiatives that are constrained by lack of funds include:

the settling of claims by the Land Commission, affecting 250,000 people and 500,000 ha of land (R503 million);13



providing a basic income grant for all South Africans (R6–7 billion per annum);14 and



providing access to clean water for eight million people in rural areas (R5.1 billion).15

A third way in which military expenditure may hinder economic growth is via debt. In the 1970s and 1980s, military imports by developing countries were the equivalent of a quarter of new foreign debt incurred. South Africa’s new arms purchases, it should be noted, are to be largely funded by increased foreign debt. The servicing of this debt (payments of principal and interest) will be listed elsewhere in the annual budget. Thus the true cost of defence will be underestimated.

Economic growth is generally accepted as making development (and the opposite side of the coin, poverty reduction) easier to achieve, but it is clearly not a sufficient condition for development. Economic growth is a means to an end—the end being development or poverty reduction. There are two main views concerning the most effective way of reducing poverty. The current orthodoxy, strongly encouraged by the World Bank and International Monetary Fund, is to encourage economic growth by minimising government involvement in the economy and maximising the role of the private sector. While the initial beneficiaries will be high-income business people, benefits are expected to permeate throughout society and thus reduce absolute poverty. The opposite view, with links to the ‘basic needs’ approach of the 1970s, is based on the belief that even rapid growth may make little or no difference to the lives of the poor. Consequently, this approach favours the direct provision of basic goods and services to low-income people. Government expenditure choices—be they of immediate consumption goods or expenditures like education with long-term investment effects—are therefore crucial.

Military expenditure works against both approaches to poverty reduction. It has a negative effect on economic growth via reduced savings and it leaves less government financial resources available to directly meet the needs of the poor. Given the extent of poverty within South Africa, milex of the order of R16 billion per annum is vastly excessive. Insofar as it hinders government from tackling poverty, it actually increases the prospects of internal insecurity.

To sum up the argument so far, evidence has been provided which suggests that South Africa’s present and planned milex is irrational in a number of respects:

It is out of line with security needs, both present and future;


it is inconsistent with stated national security policy objectives; and


it has a negative effect on economic growth and development

The paper goes on to argue that it is probably not a cost-effective way of providing for South Africa’s external security, nor a cost-effective way of achieving the other benefits listed in Table 3.

An important prior research question, therefore, is to identify the underlying determinants of South Africa’s milex. Why does such irrational behaviour persists?

What is driving South Africa’s milex?

Five broad explanations of milex have been identified:16

The past. Countries inherit a military establishment and through the annual budget-setting process, have tended to at least maintain the budget shares going to milex. That is, the best predictor of milex in any year is that of the previous year.



Geo-strategic considerations. A country’s strategic situation—involvement in armed conflict or the perceived likelihood of such involvement—is a major determinant of milex.



The politics of the budgetary decision-making process. The relative strength and ability of the military pressure group will have a strong impact on milex allocations.



Pressure from arms suppliers. The arms industry, from within or from outside a country, may in various ways press a government to allocate more to milex than it otherwise would.



Financial and economic factors. This will determine the ability of a society to allocate resources.

It is clear that the above list does not uncover the underlying motives. In particular, milex is seen as a means of acquiring status in a region and the wider world; this motive is very important to the South African government. The economist’s question, again, is how cost effective is milex as a way of achieving such status? Are there more cost effective means of increasing status?

In addition, two beliefs—almost universally accepted but open to question—underlie the milex allocation process, i.e. that a strong military is the best means of achieving security and that a strong military is a necessary part of achieving a level of national pride and status. These beliefs underpin the efforts of the military pressure group, which includes the military, related politicians and public servants and arms suppliers (domestic and/or foreign). The military pressure group basically acts so as to maximise its share of government resource allocations in a geopolitical context and subject to an economic constraint. In times of war or when a threat is perceived, the former dominates. In more normal times the size of a country’s gross national product (GNP), and the share of GNP which a government can capture, act as general limits to milex. More research needs to be done to uncover the underlying determinants of milex.

As far as the arms deal is concerned, it is widely believed that a grossly over-optimistic estimate of the spin-offs from the arms deal was prepared by the Department of Trade and Industry and this was readily accepted by Cabinet.

Alternative ways of achieving security

A major task for economists is to evaluate alternative ways of achieving policy objectives in order to determine the most cost effective. In respect of security, however, there has been an extreme emphasis on a strong military as the appropriate means and little consideration of alternatives. There are three compelling reasons for considering alternatives. The first reason lies in the changed nature of security threats. A second concerns cost effectiveness. The alternatives listed below are certainly less costly than the military although, like the military, their potential effectiveness is uncertain. On this point, military solutions frequently do not resolve conflicts. Rather they impose a result by force and often result in a large and costly military presence in the long term. Examples include the Falklands, Iraq and Kosovo. Third, there is the matter of conscience: we know in our hearts that the threat of violence, preparation for violence and violence itself are wrong and that the use of such means, even to achieve moral ends, ends up by corrupting those individuals and societies who use them.

It is essential to develop effective, non-violent ways of achieving security. The Defence Review emphasises that “the use of threat of military force [is] a measure of last resort in the face of aggression when non-violent forms of conflict resolution have failed”. This section briefly examines five such alternatives. It excludes some alternatives, which remain essentially military in character, e.g. moving to a defence-only military capability or confidence-building measures such as greater transparency concerning the nature and extent of military capabilities.

A government may choose to unilaterally dispense with its military forces, which means, among other things, that it is committing itself to deal with conflicts in non-violent ways. An impressive example in this respect is Costa Rica, which chose to demilitarise in 1948 and which, in 1983, announced its ‘permanent, active and unarmed neutrality in respect of armed conflicts elsewhere in Central America’. Despite its location—in a region with a long history of armed conflict, oppression and human rights abuses—Costa Rica has a flourishing democracy and social indicators far ahead of countries in the region and of countries elsewhere with similar levels of gross domestic product per capita. The explanation for the latter appears to be the direct application of many of the resources not spent on the military to social expenditures.17



A government may negotiate non-aggression treaties, possibly in conjunction with bilateral reductions in military strength, with one or more neighbours. Indeed, UN members are under an obligation to settle their disputes peacefully, and many countries have non-formalised but powerful understandings with their neighbours that they will not settle disputes by force. A functioning democracy can also deal with internal differences of opinion by debate and through the electoral system.18



A range of institutions can be established to facilitate the non-violent handling of disputes. These may include:

– forums where the concerns, positions and interests of disputants can be aired and clarified. These may allow for negotiation with a view to securing agreements and may defuse disputes before they reach a heightened level;

– institutions to help in the resolution of conflicts. Conflict resolution seeks a win-win outcome which is satisfying to both sides.19 On criteria such as cost, durability of the outcome and the quality of the relationship between the disputants, the attempt to reconcile the interests of disputants is superior to alternatives based on ‘rights’. A classic example is the five-year mediation carried out by the Papacy between 1979 and 1985 concerning the Argentinean–Chilean dispute over three islands in the Beagle Channel. The peaceful resolution of this conflict stands in stark contrast to the UK–Argentina war over the Falklands, despite many similarities in the two disputes; and

– when resolution cannot be achieved, institutions to adjudicate in disputes between and within countries.



Governments and non-governmental organisations can train and educate a population in non-violent conflict resolution and conflict management. The reasoning behind this suggestion is that people are currently educated, in the widest sense of the word, to deal with disputes in certain ways; often these emphasise power and rights with recourse to violence being a distinct possibility. This suggestion involves educating people in ways of resolving conflict, with the advantages mentioned above.



A population can be trained in social defence—in non-violent resistance to an invader or usurper of power—so as to make an invasion or takeover extremely costly and perhaps untenable.20 The success of informal non-violent resistance by many independence movements and by the Norwegian and Danish populations to Nazi occupation during the Second World War shows its potential effectiveness.

This does not exhaust the list of non-violent, non-military alternatives that promote security. Preventive diplomacy is, of course, widely practiced, educational and cultural exchanges between countries could be multiplied many times and development and democratisation can reduce the intensity of internal sources of insecurity. The central point of this section is that there are many relatively cheap ways of working for security and that the current focus on military-based security is almost certainly economically unjustifiable. There are also reasons, based on considerable experience, to believe that non-military alternatives are more likely to resolve conflicts and result in a sustainable peace.

Despite statements committing itself to non-violent means of resolving conflicts, the South African government puts very few resources into these, apart from preventive diplomacy, by comparison with the R16 billion spent annually on the military. Given the apparent cost effectiveness of these alternatives, this is not rational behaviour.

Is there a way forward?

In this article, the rationality of conventional ideas about security and South Africa’s defence expenditure have been questioned. Some alternatives have been suggested which will provide more security at less cost. Taken seriously, these suggestions amount to a demilitarisation of South Africa.

What would a demilitarised society look like? We cannot do better than quote from Kathleen Lonsdale, writing of Britain in the 1950s. She contends that it would be far easier to implement and monitor total disarmament than partial disarmament. In addition, non-violent methods will be much more likely to work if there is no military alternative to fall back on.

Total disarmament would mean the complete disbanding of all military organisation and the complete withdrawal from all military commitments … It would have to be accompanied by a declaration of neutrality and an offer to permit any kind of inspection within our own country. It would mean the withdrawal of spies and secret agents from other countries … We would be forced to negotiate in a spirit of mutual seeking for mutual advantage. It would give us a very real incentive to improve techniques of negotiation and conciliation in disputes between other nations …

Total disarmament would not [just] be an extreme form of partial disarmament … it would be something quite different … At present our attitude is ‘If you eat my grandmother, I’ll eat yours. But if you will agree not to eat my grandmother, I’ll agree not to eat yours either, but I will jolly well look out to see that you are not beginning to boil the water in the saucepan’. What we need to do is to develop a horror of cannibalism, a horror of the crime of war …

The really important thing is that men and women themselves should learn to abhor war and all preparations for war, not only in one country—although some country must set an example—but in every country. And that they should learn to abhor it so much that they were willing to accept the readjustments that the absence of war … and force might mean.21

Moving towards a demilitarised society involves some very challenging readjustments. A Ministry of Peace would need to be established to oversee them, some of whose central tasks would be:

educating the population in the ways of non-violent conflict resolution;



developing appropriate institutions to meet security needs and support conflict resolution efforts;



developing alternative ways of providing civil defence;



carefully and generously demobilising military forces and destroying weapons.

Violence and force are choices. We are able to choose non-violent alternatives. There is considerable evidence22 to give us confidence that we can devise and implement alternatives which will lead us towards a more just and peaceful society. Encouraging our people to accept these alternatives would ordinarily be the work of priests and poets, but we hope we have demonstrated that economists, too, have a significant contribution to make.

Notes

This article is a revised version of the author’s inaugural lecture presented at the University of Natal in September, 2000.



T Seybolt, Major armed conflicts, in Stockholm International Peace Research Institute (SIPRI), SIPRI Yearbook 1999. Armaments, Disarmament and International Security, Oxford University Press, Oxford, 2000, p 15.



United Nations Development Programme (UNDP), Human Development Report 1994, Oxford University Press, New York, 1994, ch 2.



Ibid, p 50.



J Galtung, Twenty-five years of peace research: ten challenges and some responses, Journal of Peace Research 22(2), 1985, pp 141-158.



W Eckhardt, Death by courtesy of governments, 1945–1990, Peace Research 24(2), 1992, pp 51-55.



<http://www.mil.za/secretariat/SA%20> SA needs defence force, accessed 6 March 2000.



<http://www.mil.za/secretariat/SA%20> accessed 3 March 2000.



C Conetta and C Knight, Inventing threats, Bulletin of the Atomic Scientists 54(2), 1998, p 35.



Ceasefire Anti-war News, 3/00, December 2000, p 3.



E Benoit, Growth and defense in developing countries, Economic Development and Cultural Change 26(3), 1978, pp 271-280.



For example, P Dunne, Economic effects of military expenditure in developing countries: a survey, in N Gleditsch, O Bjerkholt, A Cappelan, R Smith and P Dunne (eds), The peace dividend, Elsevier, Amsterdam, 1996, pp 439-464.



<www.bday.co.za/bday/content/direct> More money needed to avoid Zimbabwean situation, accessed 30 June 2000.



Mail & Guardian, 21-27 January 2000, p 39.



Millions have no access to clean water, Mail & Guardian 23-29 June 2000, p 42.



R West, Determinants of military expenditure in developing countries: review of academic research, in G Lamb and V Kallab (eds.), Military expenditure and economic development, World Bank Discussion Papers No. 185, 1992, pp 113-145.



G Harris, Military expenditure and social development in Costa Rica: a model for small countries? Pacifica Review 8(1), pp 93-100.



P Harris and B Reilly, Democracy and deep-rooted conflict: options for negotiators, Institute for Democracy and Electoral Assistance, Stockholm, 1998.



W Ury, J Brett and S Goldberg, Getting disputes resolved, Jossey Bass, San Francisco, 1988.



G Sharp, Deterrence and defence by non-violent sanctions, Social Alternatives 6(2), 1987, pp 9-18.



K Lonsdale, Is peace possible? Penguin, Harmondsworth, 1957.



See the references in footnotes 17 to 22 and, for a wide-ranging review, G Paige, To leap beyond yet nearer bring: from war to peace to nonviolence to nonkilling, International Journal of Peace Studies 2(1), 1997, pp 97-108.