The SIPRI Yearbook

4. The SIPRI Yearbook

I. The idea

The SIPRI Yearbook of World Armaments and Disarmament was entirely Robert Neild’s idea. Before his appointment, in all the discussions of what SIPRI should do, it had not featured. Neild put the proposals to the Governing Board in May 1967–the first meeting of the Board which he attended as Director. The Yearbook came eighth and last in the list of projects considered; it was briefly noted that the director had put the suggestion to the Board and it was approved.

The Board were clearly not going to oppose an idea of the director who had just been appointed; however, the minutes leave the clear impression that this particular project was not foremost in the Board’s mind at the time. They probably saw it as a relatively minor adjunct to proper research projects. It was certainly not seen then as something which would eventually become SIPRI’s ‘flagship’.

At the Economic Commission for Europe, Neild had worked on a yearbook which reviewed the state of the European economies–that is, of the whole of Europe, not just Western Europe. Neild was familiar with the family of yearbooks produced by United Nations agencies and used these as examples in presenting his case for the new SIPRI Yearbook:

SIPRI’s Yearbook, therefore, began with a rather different view of world security from that, for example, of the Military Balance, the annual report published by the London Institute of Strategic Studies (ISS–now the International Institute for Strategic Studies, IISS). The title of the ISS publication is indicative. It implies a ‘balance of power’ analysis. War can break out when there is an imbalance of power–as in 1939, when France and Britain failed to rearm adequately to match the growing military power of Germany. The postwar need was to apply the lessons of the pre-war period to the containment of Soviet expansionism. The main question behind the numbers in the Military Balance was this: Are NATO forces strong enough to contain those of the Soviet Union and its allies?

The world view–and the set of values–of the SIPRI Yearbook was different. The Preface to the first Yearbook says:

SIPRI was indeed committed by its Statutes to assess the consequences of world military developments for world security, not national security or NATO security. SIPRI’s first Statutes (1966) required it to recruit its researchers from different regions and from states with different political and economic systems–that is, there had to be researchers from the Soviet Union and Eastern Europe as well as from the United States and Western Europe. It was thus in the same mode as the Pugwash Conferences, which brought together scientists from the USSR as well as the USA, and which worked in the spirit of the 1955 Russell—Einstein Manifesto: ‘We are speaking on this occasion, not as members of this or that nation, continent or creed, but as human beings, members of the species Man, whose continued existence is in doubt’.

II. Armaments and conflict

There was a recurring criticism of SIPRI’s concentration on matters of arms and disarmament: that SIPRI researchers believed that armaments were the main cause of conflict. As the Yearbook Preface makes clear, SIPRI never held to this simple view. However, it did dispute the opposite view–that ‘the world war industry’ was some kind of epiphenomenon, a secondary symptom, with no causative role at all. These are some of the points made to explain why SIPRI considered that armaments and disarmament mattered.

The new factor (new in historical terms) is the vast size of peacetime military establishments and the exceptionally rapid rate of technical change in the military sector. There is a very large industrial and research infrastructure entirely devoted to the design, production and refinement of weapon systems. The scientists (and there are a great many of them) employed in military research and development examine every new scientific development for potential military applications. There are many ways in which the existence of something which can legitimately be called the world war industry changes the study of the potential causes of the use of force in international relations:

1. The military sector is not politically passive. There are, after all, a great many states where civil control of that sector is either weak or non-existent. Even in states with strong civil control, the military—industrial sector is powerful in its own right in the political system. It is naturally in the business of attempting to preserve its size and status, and indeed to provide justification for increased resources if possible. It therefore presents the picture that security can only be obtained by military superiority over potential enemies: it needs an enemy. Further, because military intelligence is part of the same complex, it is in a position, by selective use of evidence and worst-case analysis, to exaggerate the potential military threat. It is inevitably in the business of encouraging distrust of other nations.

2. The constant technological competition, in devising new weapon systems and upgrading old ones, keeps the whole structure unstable. Politicians fall very easily for the ‘fallacy of the last move’–that the adoption of this or that particular device will ensure military superiority indefinitely. They do not ask what happens when other states follow suit. Thus the USA decided to install multiple independently targeted warheads on their missiles, without asking what would happen when the USSR, with its much more powerful rockets, developed them as well.

3. The same is true of weapon deployments. When the USA decided to forward-base some of its nuclear-armed missiles in Turkey, a country adjacent to the USSR, no one asked what would happen if the USSR decided to try to do the same–as it did in Cuba. We now know that the world then came very near to a nuclear weapon exchange.

4. As weapon systems become more and more complex, the economics of their production requires arms sales abroad to keep down the unit cost. Other states are persuaded that their security, too, requires them to buy the new fighter-bomber; then neighbouring states follow. States expect some return on their heavy investment in weaponry: they can only obtain a return by using them or threatening to use them.

5. The massive postwar build-up of the nuclear weapon industry means that now the process of dismantlement poses a great threat of proliferation. With a great number of nuclear weapon technicians made redundant, and with fissile material inadequately guarded, the world will be lucky if in the future it does avoid a nuclear weapon threat from some new quarter.

These are some of the conflict-creating consequences of the world war industry. For a long period it has led to a militarization of international political relations. For 40 years military items–in the Strategic Arms Limitation Talks (SALT), the Strategic Arms Reduction Talks (START) and the Anti-Ballistic Missile (ABM) Treaty negotiations–were the dominant items on the agenda for any dialogue between the USA and the USSR. This distracted politicians from their proper business: to build a nexus of cooperative and international legal arrangements between states–such that the idea of settling disputes by military force falls off the agenda of possible action.

III. Data banks

The Yearbook had to provide a wide range of basic information about the military sector and the attempts to constrain it. This included the construction of a number of statistical data banks–although the Yearbook did provide a good deal more than a set of statistical tables. The data banks covered nuclear weapon tests, military expenditure and the arms trade. For all three, accurate figures were hard to come by. One of the long-term objectives of the Yearbook was to build up an accurate picture of the world war industry–so that one knew as much about this industry as about the world oil industry, for example. For this objective, numbers were obviously needed; without them, the picture remained anecdotal and highly uncertain. In the interwar years, the League of Nations did publish a Yearbook with some material on arms production, deployment and trade. After World War II, for 40 years, the United Nations did not publish any material of this kind; that left a large information gap.

For US and Soviet nuclear weapon tests, the US Atomic Energy Commission (AEC) provided some figures; the Soviet Union provided none. However, the AEC said that they did not report all US tests; nor did they report all the Soviet tests which they detected, since they were reluctant to reveal the detection threshold of their monitoring systems. Fortunately there was another source–the Swedish seismic detection centre at Hagfors; this reported about twice as many Soviet tests as the AEC. After 1966 France announced its tests, and China did so as well. In general, the estimates in the early Yearbooks were revised upwards in later editions, as more tests became known.

The data bank for world military expenditure began with the crude collection of numbers. A large work-sheet was established for every country, and any figures for military spending from any source were entered on it. The published budget documents were examined, and letters were sent to every state. In a great many countries some parts of military spending were not included in the defence estimates. For example, when the British Cabinet–or more precisely a few members of the British Cabinet–decided that Britain needed to have a nuclear weapon, the cost of the massive investment needed was entered in the budget of the Ministry of Works and Buildings. The single figure which the Soviet authorities gave for military spending was absurdly low. The general conclusion is that defence budgets are underestimates. The margin of error is one way: the world’s military expenditure burden is greater than the total of the published figures.

The main objective of SIPRI’s military expenditure figures was to provide estimates of the ‘opportunity cost’ of military spending–the diversion of resources from civil needs. They provided a minimum estimate of the immense scale of the resources the world devoted to preparations for mutual slaughter. They also provided a kind of early warning–if military spending is rising very sharply in some region (as it was in the Middle East), this suggests a higher risk of war (which indeed did occur). Military spending is not by itself a good measure of relative military capability–and the Yearbook criticized the way in which US estimates of Soviet military spending were used to magnify the Soviet threat. There were a number of classic examples of this.

The US Central Intelligence Agency (CIA), correctly dismissing the official Soviet defence budget figure as absurd, built up a dollar estimate by taking intelligence figures of Soviet military purchases item by item and valuing them at their dollar cost. This can produce an absurdly high figure. For instance, the Soviet Union had a huge conscript army which was paid very little: consequently it was lavish in its use of manpower–indeed, many of the conscripts were used for essentially civil purposes, such as growing food for army mess-halls. The CIA dollar estimate given to these conscripts was the average pay of the American soldier: the result was a very high dollar figure, widely used in the US presentation of the threat.

A second, more complex, example shows how the economic analysis of Soviet military expenditure figures was used to suggest a much increased threat, whereas in actuality it suggested the exact opposite. Initially, the CIA assumed that the Soviet military production sector had a much higher level of productivity and general efficiency than the civil sector. In 1976 they changed this assumption and decided that after all there was no big productivity difference. As a consequence, their estimate of the Soviet military sector’s share of Soviet national output went up from 6—8 per cent to 11—12 per cent. Note that there was no change in the CIA estimate of the size of the Soviet Union’s military effort: the change in the productivity estimate simply said that its burden, on the rest of the Soviet economy, was much greater than had previously been assumed. This clearly implies that the Soviet Union was weaker, not stronger, than previously thought.

The message that reached the public, and the legislators, was the exact opposite of this–that the CIA had doubled its estimate of Soviet military expenditure. President Richard Nixon–who may or may not have understood what the CIA had done–commented: ‘In 1976 the CIA estimates of Russian military spending for 1970—75 were doubled overnight. . . . Thanks, in part, to this intelligence blunder we will find ourselves looking down the nuclear barrel in the mid-1980s’.

The third data bank which SIPRI established was for the arms trade. It was a subject which had received a good deal of attention in the interwar years but had been largely neglected in the postwar period. Public information on the arms trade badly needed some statistical base; existing material was largely anecdotal. The choice of the arms trade for research also served to show that SIPRI was not exclusively concerned with the central confrontation between the two blocs: it was a matter of interest to Third World states–although of course it was linked to the global struggle between the two power blocs as well.

The arms trade data base was the most difficult one to establish. The published trade statistics of the main exporters were useless: typically arms exports were buried in some category such as miscellaneous metal manufactures. Trade statistics from the importing side were of no help–although they were all examined. A note on the examination of the official statistics of Bahrain reads: ‘The only possible items of military use separately categorised in Bahrain’s import figures are horses and donkeys; unfortunately it is not possible to separate out those used for military purposes’.

The only other route was to build up a register of identified transactions from other sources. This required a search through a wide range of journals–particularly those which could be regarded as the trade journals of the major arms producers. This method proved to be more productive than the SIPRI arms trade research team expected in establishing a register of the trade in major weapon systems–planes, tanks, ships and missiles. (It soon became clear that there was no hope of constructing any reasonable picture of the trade in small arms.) This system, of using unofficial sources of this kind, required a good deal of ‘reliability assessment’: the general rule followed was that there had to be at least two, and preferably three, independent reports (although it was not easy to be certain that reports were in fact independent). Detective work was needed. For example, a report appears that North Korea is exporting a particular type of tank to Iran. Does North Korea produce these tanks itself, under licence? Perhaps China is exporting these tanks, via North Korea, so as to conceal its involvement? A second report, in another journal, seems independent, in that the numbers given are different. This was the kind of assessment of data required.

The next problem was to find a good way of adding up the results. The figures given for the value of contracts were of little use–sometimes these figures were simply for the initial supply of the weapons; sometimes they covered spares, maintenance and training. Exports were often heavily subsidized–the price charged was a political decision. The purpose of the figures was to establish some estimates of the volume of the arms trade: Was it rising? Who were the main recipients, and who were the main exporters? The basic method chosen was to take the known dollar cost, in a base year, of a full set of US weapon systems, and then to match the weapon systems of other states with the US products. So, for example, a French Alouette helicopter was matched with the same generation of US helicopter which had roughly the same capabilities, and the Alouette was then given the appropriate dollar price. This was essentially a ‘purchasing-power-parity’ exercise. It was also a labour-intensive exercise: at the initial stage, for example, it involved establishing a comparable value for over 200 types of aircraft alone. The researchers had to learn a great deal about weapon systems–bringing to mind a saying of Alva Myrdal: ‘We need some peace researchers who know as much about military matters as the military do themselves’.

The result of this research work was a register of transactions in major weapon systems, which has been published in each Yearbook since the first edition, and quantitative estimates of the trends in this particular sector of the arms trade.

IV. Presentation problems

There were certain presentation problems for a Yearbook on military matters which was neither Eastern nor Western. Care had to be taken about the language used. In the West, the phrase ‘beyond the Iron Curtain’ was in common usage. To a Soviet reader, it was a provocation. So, too, was any reference to the ‘free world’. There was also a problem about the amount of published information–on the Yearbook’s subject matter, a good deal was publicly available in the West, and very little in the East. This created problems. Any discussion of the technological arms race–the development of new weapon systems–needed examples. The first Yearbook for instance, discussed the many technological advances in the development of submarine–launched ballistic missile systems. The example had to be American, since so little was published about Soviet weapon systems; indeed, almost all the examples of developing military technology had to be American for this reason. The text had to point this out frequently: the Soviet Union was also engaged in the military technological arms race, although it is true that in many developments the USA was in fact in the lead and the USSR followed.

The information about Soviet weapon developments came almost entirely from the United States. This was another problem. In assessing any source of information, this question should always be asked: ‘Do those providing the information have any reason to bias it one way or the other?’ In this case, the answer was that they did. The justification for the scale of US military expenditure was the Soviet threat: the greater the threat, the greater the justification.

There were many examples of this. To take just one: in the late 1960s the Pentagon made great play with Soviet developments in ABM systems, since they were waiting for congressional approval for funds for a US system. They said that the USSR had gone so far as to develop a ‘loitering’ ABM missile, which could be fired and then could ‘loiter’, waiting for the incoming offensive missile. After two appearances in the Pentagon’s justification for its budget requests, this loitering missile disappeared from the record.

Occasionally, in testimony to Congress, US military men could not restrain themselves from making scathing remarks about some item of Soviet weaponry. Thus a general, commenting on a new Soviet tank, said that it needed a left-handed dwarf to drive it and that the USSR would be in trouble when it ran out of left-handed dwarfs.

V. The first Yearbook

The political background

The first Yearbook–SIPRI Yearbook of World Armaments and Disarmament 1968/69–went to the typesetters in July 1969, with some amendments to bring it up to date in September.

The world background was the following. The Non-Proliferation Treaty had been negotiated and opened for signature in July 1968. However, there had been some delay in ratification in the West, because of the invasion of Czechoslovakia in August of that year and the enunciation of the Brezhnev doctrine. This was set out in Pravda in September 1968 as follows. It theoretically accepted the Khrushchev principle that each nation might take its own ‘separate road to Socialism’, but added the crucial reservation that the freedom to do this

In 1968 and 1969, US intervention in Viet Nam was at its peak. In both years there were over half a million US soldiers in Viet Nam, and in 1969 over 19 000 were killed. In the November 1968 US presidential election, Nixon had been elected. Nixon came to office with two incompatible objectives in mind: ‘I would have to act on what my conscience, my experience and my analysis told me was true about the need to keep our commitment. To abandon South Viet Nam to the Communists now would cost us inestimably in our search for a stable, structured, and lasting peace . . . I would have to end the war as quickly as was honourably possible’. He also had it in mind to practise ‘linkage’–for example, to offer the Soviet Union progress in arms control in exchange for Soviet pressure on the North Vietnamese or cooperation in the Middle East. Such linkage could only work, of course, if for some reason arms control was judged to be of more benefit to the USSR than to the USA.

Action—reaction, and threat presentation

The first Yearbook’s discussion of the nuclear weapon confrontation incorporated two themes which can be found throughout the Yearbook series: the ‘action—reaction’ phenomenon and the presentation of the threat.

The Yearbook began at a time when the United States still retained a formidable lead over the Soviet Union in strategic nuclear weapons. The USSR was beginning to catch up in the numbers of intercontinental ballistic missiles; however, it had only just begun to deploy submarine-launched ballistic missiles, and its intercontinental bomber force was only about a quarter of the size of the US fleet. How had this come about? The relevant US decisions were those of the Secretary of Defense, Robert McNamara. The first Yearbook quoted his own appraisal, because it provides a classic description of the action—reaction arms race process:

The Yearbook–written in the first half of 1969–also reports a good example of threat presentation. The Administration had changed from Johnson to Nixon in the November election, and Melvin Laird had taken over from Clark Clifford as Secretary of Defense. Clifford’s last appraisal suggested that the United States continued to have an unequivocal and considerable lead:

Laird presented a much more alarming picture of ‘the threat’. The United States had not increased the number of its launchers but had concentrated on qualitative improvements. The Soviet Union was trying to catch up from a very low level of deployment: any simple extrapolation of the rate of increase, if indefinitely continued, would obviously lead to an eventual superiority in numbers. Then the assumption is added that the USSR matches the array of US qualitative improvements. This leads to Laird’s conclusion: ‘They are going for a first strike capability. There is no question about that’.

This was another constantly recurring theme in the Yearbook’s analysis–the ‘first-strike threat’. The way in which Laird built up his picture of the threat from the USSR is a good illustration of the genre. First, it is assumed that the rate of increase in the construction of new Soviet ICBMs and ballistic missile submarines continues indefinitely. Then it is assumed that the USSR will copy all the US technological developments, including of course multiple warheads. Then there are two further assumptions. One is that the USSR will develop an effective and very extensive anti-ballistic missile system. The final assumption is that there will be some as yet wholly unidentified new development in anti-submarine warfare. Laird admitted that US submarines could not at that time be correctly located; nor was there any known technology which would enable the Soviet Union to do this in the future. But some as yet unknown new technology might at some time emerge. All this, put together, produced the threat of a Soviet first strike.

Technological competition

In a separate substantial chapter, the first Yearbook also discussed the technological nature of the arms race:

The chapter set out the figures of ‘research and development input, per unit of output’ in military procurement, compared with the average for manufacturing as a whole; for Britain, France and the USA the first figure was over 50 per cent, and the second 5 per cent. It also examined the long-term increase–from World War II to 1968–in the ‘real cost’ of certain military items, such as an air force fighter or an attack submarine, giving an average annual percentage increase in ‘product improvement’ ranging from 9 per cent to 18 per cent a year.

In the nuclear weapon field, the Yearbook discussed the main destabilizing development–the introduction of multiple warheads on the US side. It also had a long section on the extensive changes which had been made in upgrading the Polaris missile.

Multiple warheads produced a situation in which, with the regular increase in missile accuracy, a single missile on one side could threaten the destruction of a number of missiles on the other side. Whether this development had gone beyond the point of no return by 1959 is an open question: there were certainly proposals for a constraint on this development. Initially the multiple warheads were not independently targeted: they simply served to extend the area of devastation–three warheads exploding some distance from each other created more destruction than a single warhead with the same total explosive power. Later it became possible for each warhead to be independently targeted, with the possibility in some designs of up to 14 separate warheads.

The change from single to multiple warheads was an obvious major upward step in nuclear weapon technology. However, there was also a continuous process of a huge array of improvements which, cumulatively, served to produce radical changes in the capability of various nuclear weapon systems. It was not easy to convey, in the Yearbook, the wide range of technologies involved in, for example, the upgrading of the Polaris system–a matter discussed in the first Yearbook. The Polaris system was first deployed only in 1960, but by 1969 there had already been two major developments in the missile, from the
A-1 missile to the A-2, and then on to the A-3. The list of improvements is long: improved propellants, light-weight motor case materials, more advanced thrust vector control, reduction in the size and weight of the guidance system, a basic advance in the re-entry system configuration, and general miniaturization of electronic components and subsystems.

The effect of the increase in range was described in this way by the director of the Polaris programme (and quoted in the Yearbook):

Further, the development of the next stage–the Poseidon missile–was already well advanced. This represented ‘an eightfold increase in the performance of the missile’, without taking into account the multiplication of the warheads. A doubling of the accuracy leads to much more than a doubling of the effectiveness of the weapon. Accuracy is usually measured by the use of the circular error probable (CEP). A CEP of 1/4 mile means that there is a 50 per cent chance that a missile fired will fall within a radius of 1/4 mile of the centre of the target. If the CEP is halved, then the weapon yield needed to eliminate a specific target is reduced by a factor not of 2, but of 8. The Yearbook also presented a full account of the developments in chemical and biological warfare since the end of World War II. (See also chapter 5 on the history of the SIPRI CBW project.)

Military expenditure and the arms trade

The Yearbook figures showed that US military expenditure, between 1965 and 1968, had risen 40 per cent in real terms. Most of this rise was due to the Viet Nam War–although there is, as the Yearbook explained, some difficulty in calculating the incremental cost (as opposed to the full cost) of the war; the US divisions engaged in the war would have been requiring some military expenditure had they been on a peacetime footing. Soviet spending also rose over the same period by 30 per cent, according to their own figures. There were also dangerously rapid increases in military expenditure in the Middle East. The Yearbook, as the first of the series, took the opportunity for some longer-term comment on world military expenditure. In 1913, even after three years of a competitive arms race between the great powers, world military spending was only about 3—3 1/2 per cent of world output. It was roughly the same percentage in the 1930s. However, over the period from 1950 to 1968, its share in world output had more than doubled, to around 7—8 per cent. The main change was in the United States, where military spending, as a percentage of national output, had risen from around 2.5 per cent in the 1930s to around 10 per cent in the period after World War II. The Yearbook reported that ‘the world is now devoting to military purposes an amount of resources which exceeds the world’s total output in the year 1900’.

The arms trade analysis charted the spread of certain more sophisticated weapon systems–for instance, surface-to-air missiles and supersonic fighters–to Third World countries. It provided a figure for the long-term trend in exports of major weapon systems to the Third World–of 9 per cent a year (in real terms) from 1950 to 1968. It noted in particular the phenomenal rise in imports of major weapons by Middle Eastern countries–a rise of 150 per cent between 1965 and 1968. The Yearbook also published for the first time the arms trade register.

Disarmament and arms control

There was a brief summary of developments since World War II. The two main sections dealt first with the Non-Proliferation Treaty, which had been signed but had not yet entered into force; and, secondly, with material relevant to the long-delayed SALT Talks. It was argued that it was not too late to stop the deployment of MIRVs–multiple independently targeted re-entry vehicles–and the text pointed out the dangers if they were deployed. In the event, of course, they were deployed, and it was 25 years before the possibility of de-MIRVing was seriously discussed. The ENDC was discussing, in 1969, a comprehensive test ban and a cut-off in the production of fissile material. Both subjects are still on the agenda of the successor body, the Conference on Disarmament (with a membership of 61 (60 with the suspension of Yugoslavia), as of June 1996). Such is the tortoise-like approach to arms control.

The Yearbook also developed a substantial corpus of reference material: a chronology of major disarmament efforts since 1945 and a list of the states which had signed or ratified arms regulation treaties. There was also a study of accidents involving nuclear weapons and nuclear delivery systems, and a section on conflicts. This compared the data presented by 11 different conflict lists of post-World War II conflicts; it illustrated the fact that definitions of conflicts were ambiguous and estimates of their duration, and of casualties incurred, varied widely. It also provided (with maps) an inventory of world boundary disputes and presented chronologies of the Nigerian Civil War, 1967—68, and the Arab—Israeli War of June 1967. There was finally a glossary of modern warfare.

Reviews

The Yearbook was a success. Copies were sent to all delegations to the UN General Assembly, and the Yearbook was extensively quoted in disarmament discussions. It got good reviews in such prominent US newspapers as The Los Angeles Times and The Washington Post: ‘The Yearbook pulls together mountains of material previously known in detail to specialists but unavailable to the general public. It conveys a dizzying sense of an arms race spinning out of control’. It also got a good review (together with a few critical remarks) in Izvestiya.

However, perhaps the most important commendation came from Alva and Gunnar Myrdal. It was Alva Myrdal who had persuaded Prime Minister Erlander that the Swedish Government should finance an international peace research institute. Further, Alva Myrdal wanted material which could be useful to politicians struggling with arms control negotiations, and Gunnar Myrdal wanted ‘hard-boiled’ research. The Yearbook met their criteria and it had their blessing. Its success served to reinforce their view that SIPRI had chosen the right specialism.