1.1 What’s in a Name?

Augustine weapons were originally named after Norman Augustine who was the then president and chief operating officer of Martin Marietta Corporation. The company merged with the Lockheed Corporation to form Lockheed Martin in 1995. His Augustine’s Laws (Augustine, Reference Augustine1987) is a fascinating book written with a sense of humour and a tongue-in-cheek manner where the problems addressed are real and deserving of attention (Augustine, Reference Augustine1987, p. xiii). But the name has more illustrious connections.

The original Augustine brought Christianity to medieval England in the sixth century, starting in Kent and residing in Canterbury. Amongst his writings was the concept of the just war. This dealt with conditions under which the use of force is morally justified and how to conduct a war in an ethical manner. Augustine’s ideas on the just war were widely discussed. However, this Element focuses on the Norman Augustine interpretation of Augustine weapons systems (AWS) and does not consider the ideas of St Augustine.

1.2 Introduction

Norman Augustine weapons are a new class of weapons systems. They are costly, high-technology systems with costs rising substantially between generations of equipment, reflected in intergenerational cost escalation. They offer the prospect of ‘battlefield’-winning technologies and revolutionary warfare, which have great appeal to military staffs. Increasingly, capital and technology will replace military personnel. The emergence of remote warfare also raises ethical questions. Wars can be conducted remotely with operators in control centres located in different continents (Lee, Reference Lee2019). Their high and rising real unit costs also raise the future prospect that wars might become unaffordable. Augustine weapons have further implications for a nation’s armed forces and its defence industries.

Technology is an input into weapons with new technology resulting in new weapons. In turn, new weapons lead to substitution effects. History provides many examples. Guns, rifles and machine guns replaced swords, bows and arrows; castles were destroyed by cannons; radios replaced flags and pigeons; tanks replaced cavalry; helicopters are replacing tanks; steam power and then nuclear power replaced sail-powered warships; and nuclear weapons reduce the military advantage of large concentrations of conventional forces (Brauer and van Tuyll, Reference Brauer and van Tuyll2008).

Augustine weapons involve revolutionary changes in military forces and defence industries. Alongside Augustine weapons, another new technology is emerging in the form of cheap drones, which can replace manned aircraft for attack and surveillance missions. Both Augustine and drone technologies will be assessed. Three questions are asked. First, what is known about Augustine weapons and drones; second, what is not known; and third, what do we need to know for sensible debates about these new weapons? The answers to these questions are: little is known; there is a lot which is unknown; and much needs to be known.

1.3 What Are Augustine Weapons Systems?

As already stated, Augustine weapons are a new class of costly, high-technology weapons systems that are experiencing substantial cost escalation between generations of equipment. Augustine concluded that the unit cost of high-performance fighters has grown by a factor of 4.0 every ten years, with cost growth closely correlated with the passage of time (Augustine, Reference Augustine1987, p. 140). ‘New technology opens vast new capability vistas which are then crammed into each new generation of a product’ (Augustine, Reference Augustine1987, p. 140). The capability vistas are represented by information technology, electronics, computers and computer software.

Augustine weapons offer the prospect of battlefield-winning technologies. The future will continue the trend of capital and technology replacing military and civilian personnel, with personnel becoming more skilled. The high and rising unit costs of Augustine weapons raise the future prospect that wars might become unaffordable. Augustine weapons and drones have led to the emergence of remote warfare, which raises ethical questions such as whether drones operated from Europe should be used to attack buildings and human targets in the Middle East or the Far East (Lee, Reference Lee2019). These new weapons also have implications for the armed forces and the defence industries that need to be explained, evaluated and subjected to critical economic analysis, exploring the relationship between technology and weapons and their impacts on defence policy.

In one sense, Augustine weapons and drones are not new: like all weapons, they are lethal systems involving death and destruction, blood and ruins (Overy, Reference Overy2023). Also, insofar as they require human inputs for their operation, such inputs form a constraint on the operation of weapons (and all weapons). Human beings cannot fight continuously without a break; they are subject to psychiatric and emotional stress as well as physical injury from gunshot and shrapnel wounds. Studies suggest that after sixty days in combat a serviceman is no longer militarily effective (Overy, Reference Overy2023, p. 731). In future combat, artificial intelligence (e.g. robots) might remove the combat constraints of human inputs.

1.4 The Role of Economics

Economics stresses that a society’s resources are scarce and have numerous alternative uses. There is the classic ‘guns versus butter’ choice which affects all economies. More resources allocated to defence means less for civil uses: more tanks mean fewer schools and hospitals. Military forces are not immune from this simple insight. They ignore it at their peril. Generals can have the latest tanks; admirals can have the latest aircraft carriers and their jets; and air marshals can enjoy state-of-the-art fighter jets and Augustine weapons; but none of these are costless. Something has to go: just as in any competitive economy, more expensive investments probably mean job losses. Certainly, high technology is not costless. Costly Augustine weapons come at a price. Being so costly, smaller quantities are bought with impacts on the size of armed forces and defence industries, forming the ‘price’ of Augustine weapons. The long-run impact of Augustine weapons is smaller armies, navies and air forces. For example, in the 1950s and 1960s, the UK Royal Air Force had some 1,000 Hunter jet fighters; by 2023, it had just 150 Eurofighter jets. Future trends are downwards.

Cheap drones have the opposite effects. They can be bought in large quantities and used liberally: they are not too costly to use and lose! Nor do they require specialised industrial facilities: small households and unskilled labour can be used for final assembly operations. But in their current form they are not as effective as Augustine weapons.

1.5 Military Forces

At a higher and rising real unit cost, Augustine systems will be bought in smaller quantities (fewer are bought at a higher price). In contrast, unmanned aerial vehicles (UAVs) can be operated by ground-based crews; they have substantial range and endurance; and they can fly for longer periods of time compared with manned surveillance aircraft such as manned maritime patrol aircraft. But today’s drones are simple and limited in their military capabilities. For aircrews, simulators will replace actual operational flying, so reducing the wear and tear of operational flying with a reduced demand for replacement aircraft. Overall, Augustine systems need greater skills inputs for their operation, maintenance and repair, so there will be a greater demand for more highly skilled military personnel.

Drones will impact the numerical strength of the armed forces. Augustine weapons mean smaller forces, but this effect will be offset by opportunities to purchase larger quantities of cheap drones. There will be impacts on the training and support roles for military personnel. Simulators can be used for training, so reducing the need for costly manned training flights of expensive combat aircraft, with corresponding reductions in expenditure on maintenance and repair tasks.

However, cheap drones might be regarded not as substitutes for costly Augustine weapons but as complements. They might be operated jointly with Augustine weapons controlling ‘swarms’ of small drones both protecting the costly Augustine weapons and used as attack drones directed by the costly systems.

1.6 Defence Industries

Augustine weapons mean smaller volumes of costly systems for defence industries whilst drones mean the opposite, namely, larger quantities of cheap systems. Both types of weapon will mean defence industries becoming even more technology and research and development (R&D)-intensive. Augustine weapons will also create new demands for defensive systems, such as radar detection and defensive systems that are able to detect and destroy Augustine weapons (compare with the Israeli Iron Dome missile defence system).

Industry structure could be affected by a continuation of the long-run trend towards a smaller number of larger firms. National domestic weapons markets are likely to become smaller as nations find it difficult to fund costly Augustine systems, leading to incentives for international collaboration and a greater emergence of international firms. The twin technologies of Augustine systems and drones might be combined in one firm or lead to the emergence of separate specialist defence companies, much depending on the transaction costs of the alternative forms of firms.

1.7 Examples

There is considerable evidence in the UK and the USA of intergenerational cost escalation (Davies et al., Reference Davies, Eager, Maier and Penfold2011). Table 1 presents a few examples from a sample of UK fighter aircraft.Footnote ¹ More examples are presented later in this Element. Two of the examples show cost escalation factors of 3.0 or more, namely, the Meteor and the Javelin. This sample is restricted to post-World War II UK military jet fighter aircraft. For the whole period 1946 to 1959, the aggregate cost escalation factor was 4.20 in real terms, varying between 1.03 and 4.20 in real terms over a period of some thirteen years. The data provide evidence of UK cost escalation of varying magnitudes.

2.1 The Early Pioneers

Aircraft are the focus of Augustine’s Laws, so a brief history of the industry is required. Aircraft represents a new industry showing the transition from a new entrant and an infant industry to an advanced technology industry. The first manned and powered flight in December 1903 was made by the Wright Brothers (USA). The early years of the industry were dominated by pioneering efforts reflected in different designs and materials used in the construction of aircraft. The early pioneers were mostly privately financed inventors relying on income from their main business activity (e.g. motor trade), the mail business, sponsorship from wealthy individuals, income from air shows, passenger rides, patents and prizes from competitions. The early pioneers built and tested their own new designs, where testing was not without considerable risks involving injuries and deaths for the pioneer pilots. There were also various ‘races to be first’ offering prize money (e.g. first to fly across the English Channel).

2.2 The Role of Government and Wars

Governments have been central to understanding the development of the aircraft industry through their demands for military and civil aircraft. For military aircraft, governments are major or monopsony buyers of military equipment and they can use their buying power to determine the size, structure, conduct, performance and ownership of aircraft and defence industries. With civil aircraft, governments have funded development and purchased aircraft for national airlines and their purchases have further affected industry size and structure.

Wars lead to a major increase in the demand for military aircraft and their quality is reflected in their technical performance characteristics. These characteristics comprise speed, range, altitude and weapons capabilities. Wars mean demands for aircraft that are faster and can fly farther and higher; there is an emphasis on volume production and new construction techniques emerging, as well as new materials, new designs and more powerful engines. These changes are reflected in the labour requirements for the aircraft industry in the form of numbers of personnel required and their skill levels. Wars also lead to new types of aircraft (e.g. bomber aircraft in World War I). Some new technologies are revolutionary, such as the jet engine and the rocket engines for the first generation of cruise and ballistic missiles (e.g. German V-1 and V-2 rockets built in 1941–42). Further revolutionary change occurred with the development of nuclear weapons (e.g. used against Japan in 1945) and the means of delivery (e.g. missiles and rockets).

2.3 The Inter-war Years

The ends of wars lead to opposite effects to the starts of wars, with disarmament resulting in reductions in demand, plant closures, mergers, exits from the industry, diversification and job losses. After World War I, the aircraft industry had to survive through efforts to develop civil aircraft. In the 1920s and early 1930s, the civil aircraft industry was an infant industry. Major developments occurred in the USA with competition between US airlines promoting innovation in civil aircraft. An example was the introduction of the Douglas DC-3 in 1936, which revolutionised air transport with its speed and range.

2.4 World War II

World War II provided further stimulus to the aircraft industry, affecting both the quantity and the quality of aircraft. Monoplanes replaced biplanes; aluminium replaced wood and fabric (the de Havilland Mosquito was an exception); propellers became more efficient; undercarriages were retractable; and cockpits were enclosed. Imaginative designers innovated with advanced types, sometimes launched as private ventures. World War II also introduced the jet and rocket engines, which revolutionised both military and civil aircraft. Other revolutionary changes included the development and use of the first atomic bomb, the development of radar and electronic warfare (e.g. as in the strategic bombing of Germany in 1940–45) and the emergence of aircraft carriers and carrier-borne aircraft as new weapons of war. World War II also led to the emergence of mass production, enabling aircraft plants to achieve economies of scale and learning, and to replicate the mass production methods of the motor car industry. Here, the US aircraft industry had an impressive record. For example, it built over 15,000 Mustangs and 12,726 B-17 bombers and a total of almost 300,000 military aircraft in World War II.

By the end of World War II, the US aircraft industry was a world leader in civil aircraft, with an oligopoly industry dominated by Boeing, Convair, Douglas and Lockheed. The emergence of the Cold War led to an arms race in nuclear and conventional equipment between the USA and the USSR. Technical change was affecting the economics of the aircraft industry. Both military and civil aircraft were becoming technically more complex, costlier and taking longer to develop, leading to fewer being bought, making it more difficult to achieve volume production, especially in small domestic markets such as the European markets. Solutions to smaller volumes and rising unit costs included importing foreign aircraft and international collaboration.

2.5 Technical Progress

Technical progress led to major substitution effects with fighter and bomber aircraft replaced by missiles and rockets and atomic weapons replacing the traditional military advantage of large-scale conventional forces. Combat aircraft, missiles and rockets required greater electronics inputs, which led to the development of the defence electronics industry. Some aircraft firms acquired defence electronics businesses. Technical progress meant rising unit costs of military aircraft, resulting in industrial restructuring reflected in mergers, capacity reductions, diversification and exits from the industry.

Another example of technical progress concerns bombing by aircraft. Initially, during World War I, aerial bombing was undertaken by hand with pilots aiming by eyesight and dropping bombs from their aircraft. By World War II the technology had advanced and bombs were carried by aircraft and dropped by electrical methods initiated by aircrew (specialist bomb aimers). But this method was inaccurate and was improved to allow radar to guide accuracy. Manned bomber aircraft were replaced by first-generation unmanned cruise missiles (V-1 rockets, March 1942) and first-generation unmanned ballistic rockets (V-2 rockets, December 1942). By 2024, bombing was automated and controlled by computers and electronics in extremely accurate missiles and rockets where pilots and specialist bomb aimers were no longer required.

After World War II, the jet engine revolutionised commercial air travel. The British de Havilland Comet jet airliner entered service in 1952 but suffered technical failures which almost bankrupted de Havilland. The US Boeing company developed a rival, the Boeing 707, which entered service in 1958. A new entrant to the civil aircraft industry emerged in 1970, namely, Airbus Industrie, which became a major rival for the US civil aircraft firms. Collaboration led to longer production runs, enabling European firms to compete with their US rivals. By the late 1990s, Airbus was a duopoly with Boeing in the world market for large civil jet airliners. Inevitably, controversy arose over the financing of Airbus aircraft. Boeing accused Airbus of receiving illegal subsidies in the form of launch aid, with Airbus responding that Boeing received subsidies from military and research contracts and state and local subsidies. In 2019, the World Trade Organization (WTO) ruled that the EU illegally provided support for Airbus and in 2020 the WTO declared that US benefits to Boeing violated trade rules. By 2021, the trade dispute between the EU and the USA over aircraft subsidies ended with a five-year accord to make sure that R&D support did not harm the other side with a move to a level-playing-field approach. Other nations became new entrants to the aerospace industry, including Brazil, China, Japan, India and South Korea. New entry meant more competition and innovation.

The next major technical change was the industry’s move into space flight. In the meantime, the aircraft industry became known as the aerospace industry. The space race between the USSR and the USA started in October 1957 with the launch of satellites (e.g. Sputnik). The first crewed spaceflight was achieved by the USSR in 1961 and the USA landed a man on the moon in July 1969. The time-period from the first manned flight in 1903 to the moon landing in 1969 was less than seventy years, which is a further example of the industry’s rate of technical progress. More recent developments have seen the privatisation of space involving the private finance and provision of both launchers and commercial space travel (e.g. Space X (Musk) and Virgin Galactica (Branson)). Privatisation meant that private finance replaced the industry’s traditional reliance on state finance. One result was that wealthy individuals had a major role in the privatisation of space. Future developments include commercial space flights and greater militarisation of space. Overall, technical change has led to AWS (Augustine Weapons Systems), which are the focus of this Element.

3.1 Augustine’s Chapter 16

The starting point is Augustine’s chapter 16 dealing with the high cost of buying, which resulted in the most famous of Augustine’s Laws (Law XVI: Augustine, Reference Augustine1987). Law XVI states: ‘In the year 2054, the entire defense budget will purchase just one aircraft. This aircraft will have to be shared by the Air Force and Navy … except for the leap year, when it will be made available to the Marines for the extra day’ (Augustine, Reference Augustine1987, p. 143). This law was derived by projecting forwards in time the trend curves for the US national defence budget and the unit cost of tactical aircraft, which showed that in the not too distant future these curves intersect. This observation led to the formulation of what is known as the First Law of Impending Doom or the Final Law of Disarmament (Augustine, Reference Augustine1987, p. 143). Some commentators have suggested that the eventual outcome by the late twenty-first century will be one grand military vehicle – the Battlestar Galactica. An example of extrapolation at its best! Predictions have been made that the eventual outcome will be a single-tank army, a single-ship navy and Starship Enterprise for the air force! Applied to the UK, it was predicted that the UK will reach the single aircraft situation two years before the USA (Augustine, Reference Augustine1987, p. 144).

The high cost of buying chapter states that ‘the unit cost of certain high-technology hardware is increasing at an exponential rate with time’ (Augustine, Reference Augustine1987, p. 140). The point is illustrated with the example of fighter/tactical aircraft where the costs of each aircraft have grown by a factor of 4.0 every ten years, with unit costs closely correlated with time rather than with changes in speed, weight, maneuverability or other technical performance parameters. Similar trends apply to bomber aircraft. There is no ceiling or end in sight (Augustine, Reference Augustine1987, p. 141, fig. 19).

3.2 The High Cost of Buying Today

The same trend of rising unit costs applies to ships and tanks where their generally lower technology levels lead to lower unit cost growth by a factor of 2.0 every ten years. Rising unit costs over time of a similar magnitude are also apparent for commercial aircraft, helicopters, motor cars and houses, as well as ‘certain other commercial products’ (Augustine, Reference Augustine1987, p. 140). The cause of rising unit costs is that ‘new technology opens vast new capability vistas which are then crammed into each new generation of a product’ (Augustine, Reference Augustine1987, p. 140).

The electronic content of aircraft has been a source of rising unit costs, with the number of electronic component parts growing by a factor of 2.0 every ten years. Whilst the number of parts has increased, their capabilities have also increased enormously (Augustine, Reference Augustine1987, p. 125). There followed Augustine’s Law XV: ‘The last 10 per cent of performance generates one-third of the cost and two-thirds of the problems’ (Augustine, Reference Augustine1987, p. 138). Next in the development of new technology is software, which was almost non-existent a few decades ago but has become the dominant element in the design of most major high-tech systems. The result is another of Augustine’s Laws, Law XVII: ‘Software is like entropy. It is difficult to grasp, weighs nothing, and obeys the Second Law of thermodynamics; i.e., it always increases’ (Augustine, Reference Augustine1987, p. 152). Studies suggest that over the last twenty-five years, the hardware:software proportions of major systems costs have shifted from an initial eighty:twenty hardware:software ratio to a future ratio approaching twenty:eighty and that software grows by a factor of 10.0 every ten years. Augustine went further, suggesting that as the cost of various items within a family of products increases, reliability does not improve; instead, it worsens (Augustine, Reference Augustine1987, p. 158). He also had views on pressure groups: who were they and what were their specific objectives? His answer was ‘no problem’; the group will find you and they will be vociferous about what they want, which is usually money. And there is valuable advice on the size of committees: the optimum committee has no members!

To summarise, the trends over time are for major weapons such as aircraft, warships and tanks to become more advanced technically; they are costly and becoming costlier. Rising costs mean smaller numbers with forecasts of a single-ship navy, a single-tank army and Starship Enterprise or Battlestar Galactica for the air force! The prediction goes even further with forecasts that by the twenty-second century there will be one single grand fighting machine encompassing a whole nation or perhaps a whole planet! There is, though, a contrary trend reflected in the development of small cheap drones, capable of undertaking some of the missions of costly Augustine weapons.

4.1 Lancaster’s Approach

There are two further approaches to analysing Augustine weapons, namely, Lancaster’s theory of demand and a real options approach. Each provides further insights into understanding Augustine weapons. Lancaster’s novel contribution focused on goods being demanded not as goods but for their characteristics. For example, a cruise missile offers speed, range, lethality and accuracy in locating a target. Similarly, a fighter aircraft provides speed, duration, altitude and target identification. Some of these performance features might be possible substitutes, whilst others might be complements, with the whole package defined as a bundle of characteristics (Markowski et al., Reference Markowski, Brauer and Hartley2023). Different characteristics will be attractive to each of the armed forces and they will be combined into Augustine weapons offering their users ‘good’ value for money and ever more complex and costlier weapons.

4.2 Real Options Approach

A real options perspective can be used to analyse Augustine weapons. The design and development phase of a new weapon can be viewed as a sequence of real options. All weapons offer a range of design and development options. At the start, initial design work is required, followed by different stages of development leading to a technology demonstrator which, for aircraft, might offer prospects of flying before you buy. Each stage of development represents an option that might or might not be pursued, in such forms as modifications and product enhancements. For example, the Ministry of Defence (MoD) requires some new equipment that is not available off-the-shelf. As an alternative to offering a cost-plus or fixed price contract for development, the MoD could announce its willingness to buy call options from firms giving it the right to buy a specific number of units at a fixed price at some future date. There could then be a competition inviting bids and selecting the firm offering the best value for money option contract (Weston, Reference Weston1996). This is not to imply that the use of real options methodology in the evaluation of procurement choices is simple and could easily be adopted. A major problem with defence investments is evaluating the gains: there are no measures of the value of defence outputs and usually it is assumed that inputs equal outputs, which is most unsatisfactory.

5.1 The Main Drivers

The MoD survey identified the main drivers of IGCE as economy-driven factors and customer-driven factors. Economy-driven factors are those over which the government has little control, such as wage rates, equipment costs and other input costs. Customer-driven factors are ones where the government has more control, including equipment complexity reflected in performance characteristics arising from military threats, over-specification of proposals (gold plating), government regulatory requirements and declining production volumes.

5.2 Why Are Defence Goods More Expensive?

The reason why defence goods are more expensive than other goods is partly owing to monopoly and the associated monopoly inefficiency as well as small production numbers and export restrictions. Other factors contributing to IGCE include:

1. 1. optimism bias, which is the tendency of appraisers to be optimistically biased about cost estimates;

2. 2. poor understanding of risks and assumptions such as erroneous assumptions about the increased probability of cost growth;

3. 3. changes in operational capability;

4. 4. reduced competition in the UK defence market and the reduced number of suppliers;

5. 5. use of non-competitive cost-based contracts;

6. 6. procurement policy and incentives to develop new, less costly capabilities – often, the UK develops bespoke equipment rather than buying off-the-shelf from the world market;

7. 7. moral hazard – human behaviour means that contractors have incentives to under-estimate costs and over-estimate benefits to win contracts;

8. 8. workforce challenges – for instance, large projects may require the rapid build-up of the workforce, using inexperienced workers (who are less productive) and hence incurring higher cost.

5.3 Solutions

The survey proposed various solutions for cost escalation. These included improved procurement stability, with MoD publishing its long-term procurement plans and longer-term contracts. There were proposals for increased competition and more international cooperation. Also, it proposed addressing skill shortages, taking a new approach to industrial strategy, maintaining value for money in the absence of competition (e.g. via the Single Source Regulations Office (SSRO)) and increasing competition where appropriate. These proposals resemble a ‘wish list’ which is strong on ‘buzz words’ and demonstrates apparently great activity but lacks analysis and delivery. There remains a need for an economic model of cost escalation. Such a model would start from a trade-off between development time and development costs. For a given project quality (performance), a shorter development time is costlier: for example, you can land a man on the moon sooner, but at the expense of higher costs. This model predicts that cost escalation results from faster development (a time–cost trade-off) or from a choice of a more advanced project or from deliberate under-estimation of costs at the start of a project (Tisdell and Hartley, Reference Tisdell and Hartley2008, p. 384).

6.1 Identifying Firms Supplying Augustine Weapons

Table 2 presents potential suppliers of AWS. There is no published list of such firms, so Table 2 is a starting point. The list of potential suppliers was adapted from the SIPRI (Stockholm International Peace Research Institute) top 100 arms producers based on size of arms sales and shows the top twenty-five firms. Adapting the list from the SIPRI top 100 consisted of selecting those arms firms where 50 per cent or more of their sales were arms sales. This is an arbitrary cut-off point and reflects the view that AWS firms are large firms that are defence-intensive. Not all of the firms listed in Table 2 are potential suppliers; similarly, firms outside the list might be potential suppliers. To reiterate, Table 2 is only a starting point, presenting a list of firms on a subject where there are no published data.

Some potential suppliers of AWS might not yet exist. These are firms which are possible future suppliers. The space industry provides examples. In recent years, this industry has been characterised by new entrants in the form of private space companies. They include firms such as Space X (Elon Musk), Ispace, Blue Origin, Axiomatic Space and Virgin Galactic. Existing firms such as Boeing, Lockheed Martin and Nothrop Grumann are also involved in space. Foreign companies from China and India have entered the industry. The implication is that there is scope for new firms to enter the AWS industry and become new suppliers: we do not know who they are or which activities they might discover, but private markets will seek out potentially profitable opportunities.

6.2 An Alternative Method

An alternative method of identifying AWS firms comprises identifying firms that supply advanced technology and costly weapons: the weapons approach. Examples include combat aircraft, missiles, rockets, aircraft carriers, nuclear-powered submarines, anti-submarine frigates (complex warships) and tanks produced in the USA, Europe and the rest of the world. Representative projects include the US F-22 and F-35 combat aircraft, the Eurofighter Typhoon, the French Rafale, the UK ballistic missile and attack submarines, together with the UK Challenger 2 tank, the French Leclerc tank, the German Leopard tank and the US Abrams tank. The relevant data are shown in Table 3, which identifies Augustine weapons and their prime contractors, costs and quantities.

6.3 Augustine and Procurement Policy

Augustine weapons raise new issues for procurement policy involving the choice of contracts, supporting the defence industrial base and budget funding. For example, the US B-21 bomber involved a competition between two rivals, namely, Northrop Grumman and a joint team of Boeing and Lockheed Martin, with the former winning the competition. Questions arise as to how the USA will maintain its bomber industrial base to allow rivalry for any future bomber projects and the costs of such a policy (i.e. in terms of the sacrifice of other defence equipment?). Also, the current bomber project is being financed by a cost-plus incentive fee contract followed by a firm fixed price contract (CRS, 2021). Questions arise as to which type of contract will lead to the lowest acquisition price: a firm fixed price or a cost-plus contract? The literature suggests that cost-plus contracts promote cost increases and delays. But the alternative of a fixed price contract also has its flaws (e.g. it might involve firm bankruptcy and bail-outs).

The Virginia nuclear attack submarines raise similar questions about maintaining the US submarine industrial base. An annual procurement of two submarines supports two yards at the General Dynamics/Electric Boat Company and at Huntington/Newport News. Both companies have a shared industrial arrangement aimed at supporting both yards with alternate work on the reactor plants and final assembly. This two-shipyard strategy is costlier than a single-yard strategy, but it provides insurance against any yard failure. The strategy also retains key construction skills at both US shipyards (CRS, 2023). But the key issue is what are the extra costs of the two-yard strategy and will the resulting benefits make it worthwhile? Finally, questions also arise as to whether such costly projects should be funded by the service acquiring the project or by a special strategic forces fund where costs are distributed across all users and not restricted to a single service.

7.1 The Military–Industrial–Political Complex

The MIPC comprises interest groups of bureaucracies in the form of the armed forces and defence ministries, together with producer groups of defence contractors, each group pursuing its own objectives. The various groups include industrial scientists pursuing R&D activities, trade unions protecting the interests of their workers, together with local and regional associations representing local and regional interests. An international dimension emerges from international military alliances such as NATO (North Atlantic Treaty Organization). These various interest groups can be represented as agents in a principal agent framework. Agents are appointed by principals to achieve the aims of the principals. For example, in a profit-maximising firm, the owners are the principals and they appoint managers as their agents to pursue maximum profits. A problem arises since owners cannot be confident that managers might not pursue other objectives than maximum profits and might instead aim to maximise their own utility. Managerial utility comes in many forms, such as a ‘quiet life’ through on-the-job leisure or ‘empire building’ by hiring lots of attractive female secretaries or building brand new offices which give satisfaction to managers rather than the owners! On this basis, Augustine weapons are the result of agent behaviour in the procurement system.

Bureaucracies in the form of the armed forces and defence ministries act as budget-maximisers, which involves exaggerating the demand for their preferred weapons systems and under-estimating their costs. Defence contractors form producer groups which act as rent-seekers reinforcing the budget-maximising aims of the bureaucratic groups. Producer groups will also have strong preferences for profitable, cost-plus contracts rather than the incentive-based fixed price contracts. The language resulting from the behaviour of agents is predictable. A weapon must be purchased because it is ‘vital for national security’ and it will provide advanced technology for maintaining the technological competitiveness of the economy – and, of course, it provides ‘lots of highly paid jobs essential for maintaining future living standards’. Inevitably, economists are critical of such arguments. They focus on costs in terms of the alternative use value of resources: would the resources used in developing the weapons system make a greater contribution to jobs, technology and exports if used elsewhere in the economy? Opponents of such questions often demand that the alternatives be specified and are critical when the alternatives are not identified. But economists respond by explaining that market economies will re-allocate resources to their next best uses reflected in market prices (e.g. such alternative industries as construction, aerospace, motor cars, engineering, pharmaceuticals and hospitality).

7.2 The Civilian Sector

Comparisons with the civilian sector are useful for assessing AWS and suggesting alternative explanations of these weapons systems. Technical progress in the civilian sector has resulted in many products that have become smaller, faster and newer, such as mobile phones, personal computers, TVs and motor cars. Homes are now examples of high-technology hubs: they provide food, entertainment and accommodation in air-conditioned environments. Significantly, the new-technology civilian products are often supplied at lower prices compared with Augustine weapons, which are costlier with each new technology and each new type. Specifically, Augustine forecast that rising unit costs of defence equipment would result in a single-aircraft future. Why does this happen? Why are the unit cost trends so radically different between civil products and defence equipment?

Standard economic analysis suggests various differences between price determination in defence and civilian industries. Possibilities include monopoly and oligopoly in defence industries and competition in civilian industries. These differences might be reinforced where defence markets are protected markets in which national firms are awarded national defence contracts. Or, defence equipment is a tournament good where the need to maintain military superiority requires equipment which is technically superior to that of potential rivals (Davies et al., Reference Davies, Eager, Maier and Penfold2011). Or, defence procurement contracts are cost-based with soft budget constraints compared with civilian contracts, which are fixed price with hard budget constraints.

8.1 Overview

The data include some world-famous aircraft such as the Hurricane and Spitfire fighters, the Halifax and Lancaster bombers and the Mosquito from World War II. Post-1945 UK aircraft include the Meteor and the Vampire, the Hunter, Javelin and Lightning fighters, the Canberra and the V-bombers. The data available are from two distinct periods of war and peace. They also embrace major changes in technology with the introduction of the jet engine and the greater use of avionics. The data were extensively corrected by the author, so creating a unique and original data set. Corrections were made to ensure consistency and accurate costings, and data were converted to constant prices based on the UK Retail Price Index (based on 1987 = 100).

Actual contracts were available for the complete range of UK military aircraft comprising fighters and bombers of various types, trainers, transport and specialist aircraft, including air force and naval aircraft. During this period, aircraft firms emerged as specialists in fighters, bombers, trainers, and land and naval aircraft. Later, mergers in the UK aircraft industry reduced the number of firms and ended such product specialisation. The outcome by 2024 was a UK aerospace industry comprising a monopoly supplier of aircraft (BAE Systems), a monopoly supplier of aero-engines (Rolls-Royce), a monopoly supplier of missiles (MBDA) and a number of parts and components suppliers (e.g. Martin Baker ejector seats; PACE aerospace components; Marshall Aerospace aircraft maintenance, modifications and design). A number of UK aerospace suppliers were acquired by foreign firms. Examples included Dowty landing gear acquired by Safran (France) and Smiths Aerospace acquired by General Electric (USA).

8.2 Data Limitations

The data had its limitations. It reported only airframe costs and not aircraft costs, and reporting was not uniform: some contracts had missing data and not all contracts included the costs of jigs and tooling. Information was presented in current prices and was converted to constant prices with the details explained in the table notes. The information is also dated based on the period 1934 to 1964: nonetheless, these are original data not previously published, and so they provide valuable insights into the pricing and cost side of the military sector of the UK aircraft industry. Interestingly, work on the original learning curve is historical, was first published in 1936 and was based on studies which first started in 1922 (Wright, Reference Wright1936). The data allowed estimation of cost escalation for different types of aircraft and the shape of aircraft unit cost curves. In some cases, entry costs were estimated where there were alternative suppliers so that the unit costs of aircraft could be compared (e.g. where two or more firms produced the same aircraft).

9.1 Technical Change

Aircraft are a classic example of major technical change, starting with their very existence reflected in manned flight. But aircraft are not unique. Over time, most products have improved to become better, as shown, for example, by household goods and services. Consider the range of services now available within the household: communications (e.g. mobile phones), entertainment and leisure (films, TVs, sports), heating and eating (central heating, microwave ovens) and washing machines replacing laundries and the washing of clothes by hand. These changes have resulted in factor substitutions, with capital and technology replacing labour (e.g. a reduced demand for household servants).

Motor cars are a further example as they have progressed from early vehicles to today’s complex machines, which are larger, faster and more efficient with engine technology shifting from petrol to electricity power. Airliners are another example of technical progress as they have developed to larger and faster aeroplanes, capable of flying greater distances. Space travel is the next ‘new’ frontier. Overall, in the civil sector, new technology usually leads to better and cheaper products. But defence is different: new technology is costlier. Explanations for this defence feature include defence as a tournament good, monopoly defence markets and aspects of procurement policies.

9.2 UK and US Studies

There is published evidence of intergenerational cost escalation for UK and US defence equipment. Real unit cost escalation of some 8.3 per cent annually was estimated for UK combat aircraft (Kirkpatrick and Pugh, Reference Kirkpatrick and Pugh1983). A detailed study of UK defence equipment estimated real unit cost escalation at varying levels: 5.9 per cent for main battle tanks, 5.8 per cent for combat aircraft, 4.3 per cent for naval frigates and 3.8 per cent for aircraft carriers (Davies et al., Reference Davies, Eager, Maier and Penfold2011). A US study found annual real unit cost escalation rates of 6.2 per cent for surface combat ships, 5.3 per cent for attack submarines and 2.9 per cent for nuclear aircraft carriers with an overall annual cost escalation rate of 5.2 per cent for US naval equipment (Arena et al., Reference Arena, Blickstein, Younossi and Grammich2006). Evidence from the USA shows that cost escalation on fighter jets was primarily caused (some two-thirds) by customer-driven factors in the form of equipment performance (vastly increased system complexity), enhanced features and higher quality, rather than by customer-driven factors (one-third) such as inflation and higher input prices (Arena et al., Reference Arena, Younossi, Brancato, Blickstein and Grammich2008). A further study by Johnstone (Reference Johnstone2020) re-examined the fundamental Augustine Laws to assess whether they are still applicable. It found that the Law forecasting that by 2054 the rising price of tactical military aircraft would exceed the US defence budget was ‘clearly false’ (Johnstone, Reference Johnstone2020, p. 19). A similar conclusion was reached for civil airliners where cost escalation was estimated to be about half the cost growth in military fighter aircraft.

A Norwegian study examined cost escalation using controls for quality (e.g. speed, weight) and total production. On this basis, it found that the intergenerational cost escalation for helicopters, frigates, main battle tanks and fast patrol craft was close to zero (estimated at 1 per cent: Hove and Lillekvelland, Reference Hove, Lillekvelland, Hartley and Solomon2017). A Swedish study based on Swedish defence equipment estimated cost escalation as varying: 7.6 per cent for infantry fighting vehicles, about 7 per cent for fighter jets, corvettes, helicopters and 0.7 per cent for battle tanks (Nordlund, 2017). International comparisons of cost escalation for similar types of equipment found that the Swedish figures were slightly higher than the international results, probably reflecting Sweden’s traditional position as a small neutral country developing its own defence equipment (Amann, Reference Amann2022; Hartley and Solomon, 2017). Whilst there is general support for intergenerational cost escalation, the different studies have their limitations. They use different definitions of cost escalation, different time periods and different samples of defence equipment; further, equipment differs between nations and some studies are in nominal and others in real terms. Also, estimating models differ between studies and there are confusions between defence inflation and cost escalation (Hartley, Reference Hartley, Hartley and Solomon2017). All of which leads to the perennial call for more data and more research!

9.3 Learning Curves

For the analysis of trends in aircraft unit costs, there is one common feature, namely, learning curves. These show a relationship between unit costs and quantity, with unit costs declining with increasing quantity. Labour is central to the learning curve which forecasts that unit labour costs decline with greater cumulative output. This reflects workers learning by experience or by ‘doing’: the more they repeat a task, the more efficient they become, particularly in assembly operations. The traditional learning curve was an 80 per cent curve predicting that direct labour input costs declined by 20 per cent for every doubling of cumulative output (Wright, Reference Wright1936). Reality is different and actual learning curves differ from the 80 per cent curve. Increasingly, labour-intensive operations have become capital-intensive. For example, labour-intensive riveting has been automated with computer-controlled operations replacing manual labour. Critics also question the underlying theoretical basis of learning curves, arguing that ‘the phenomenon is largely a creature of empiricism supported by a few theoretical constructs’ (Baloff, Reference Baloff1966, p. 282).

10.1 Data and Sample

Data were obtained from actual UK procurement contracts for military aircraft purchased over the period 1934 to 1964. Military aircraft types varied amongst fighters, bombers, naval aircraft, trainers, transports and reconnaissance aircraft. Contracts showed the name and type of aircraft purchased, its weight, the name of the supplying firm, the dates of acquisition and delivery and the quantities purchased. Prices paid were based on unit production costs or average costs, which comprised labour and overheads, raw materials, bought-out parts and equipment, subcontracts, jigs and tools and flight testing. Unit prices consisted of average costs plus a profit margin.Footnote ² Data availability varied between aircraft types with some contracts showing unit costs with and without the costs of jigs and tools, where jigs and tools costs might be some 10 per cent of unit prices. For some aircraft, unit cost data were available for alternative suppliers (i.e. where two to three firms manufactured the same type of aircraft), showing different overhead recovery rates between different plants. The sample for detailed analysis was selected to include pre-World War II, World War II and post-1945 aircraft. The task is to define and identify Augustine weapons. Augustine asserted that from

For this study, a cost escalation factor of 4.0 every ten years is used as the strict definition of AWS, but other less strict definitions will also be used.

10.2 Case Studies

Table 4 shows trends in the levels and composition of unit cost data for UK fighter aircraft between the 1934 Gauntlet biplane and the 1959 Lightning jet combat aircraft. These examples allow assessment of cost escalation between generations of fighter aircraft and of any trends in the composition of costs. Augustine technical progress is expected to lead to changes in factor inputs between generations of aircraft. For example, higher-technology aircraft will require greater inputs of highly skilled labour. However, Augustine did not provide predictions on how the composition of unit costs was expected to vary between different types of aircraft. In the absence of such predictions, actual contract cost data are used to identify trends in cost composition for Augustine weapons (Sturmey, Reference Sturmey1964). The research challenge is to identify AWS from the available contract cost data. Here, the research question is whether all combat aircraft are Augustine weapons?

Table 4 shows cost, price and profit data for UK fighter aircraft from 1934 to 1959. Between the Gauntlet and the Gladiator, real unit costs rose little, by a factor of 1.3 (gap of forty-one months); between the Gladiator and the Spitfire, real unit costs almost doubled (two-year gap) and between the Spitfire and the Meteor (seventy-five-month gap), real unit costs rose fourfold; between the Meteor and the Hunter, real unit costs rose very little (almost a ten-year gap); and from the Hunter to the Javelin involved a more than threefold rise in real unit costs (a sixteen-month gap), whilst the Javelin to the Lightning involved only a 1.26 rise for real unit costs (a twenty-seven-month gap). Overall, the data confirm intergenerational cost escalation but at varying rates. The data on shares suggest that after the Spitfire in 1939 came rising trends in shares for labour and overheads and falling shares for bought-out parts; the Javelin shares data appear to be an outlier.Footnote ³ Unit profit rates were higher for 1930s aircraft and lower for post-1945 aircraft.

Augustine forecast that costs would rise by a factor of 4.0 every ten years, but it was unclear whether the rise was in current or constant prices. The analysis in the previous paragraph assumed that costs would rise in constant prices. Table 4 is in both current and constant prices, allowing the hypothesis to be tested in both forms. Assume that the rise is in current prices. Ten-year periods are not available so approximations have to be used. Consider the Gladiator to the Meteor, which is almost ten years: the cost escalation factor in current prices was about 10.0 based on current prices in unit production costs or unit prices. For the next generation of fighters from the Meteor to the Hunter, giving a nine and a half-year gap, the cost escalation factor in current prices was under 2.0, well below Augustine’s factor of 4.0. In total, the limited evidence gave some support for Augustine’s strict hypothesis of costs rising by a factor of 4.0 in current prices.

Table 5(a) shows cost escalation for seven UK fighter aircraft. All the fighter aircraft in Table 5(a) showed varying degrees of cost escalation in current and constant prices. Between the Gauntlet and the Lightning combat aircraft over the period 1934 to 1959 real unit production costs rose by almost forty-fold. There were substantial variations in cost escalation between each successive generation and only the Meteor conformed to the strict Augustine prediction of unit costs rising by a factor of 4.0 every ten years.

Table 5(b) shows cost escalation for UK bomber aircraft. Most aircraft showed positive cost escalation, but there were some examples of unit cost reductions. The real cost escalation factor between the Canberra and the Mosquito, as light bombers, was over 3.0 and between the Lancaster and the Vulcan, both heavy bombers, it was almost 9.0. Comparing cost escalation for UK fighter and bomber aircraft (Tables 5(a) and 5(b)), the bomber aircraft showed some cost decreases and also showed some substantial real unit cost increases. Between the Hind and the Valiant, real unit costs increased by a factor of 100.0.

Overall, real cost escalation was observed for both fighter and bomber aircraft, with a median escalation factor of 1.3 for fighters and a higher median factor of 1.7 for bombers. The variation was much greater for bombers, with escalation factors of 5.9 and 10.1. Both propellor and jet aircraft displayed cost escalation and for fighters, cost escalation increased markedly with the shift from propellor to jet aircraft (i.e. Spitfire to Meteor: Hartley, Reference Hartley2023; Hartley and Sandler, Reference Hartley and Sandler2001).

10.3 Case Study of Hawker Aircraft: 1938 to 1955

Data were obtained on Hawker fighter aircraft for the period 1938 to 1955. The data comprised all Hawker fighter aircraft over this period so that detailed study could be undertaken for a ‘family’ of aircraft, all produced by one company, embracing propellor and jet engines. For the complete period 1938 to 1955, the total real cost escalation factor was 6.1, varying between a negative value and a factor of 3.2. The results are shown in Table 6.

The problem in comparing studies is that they are for different periods and embrace different types of aircraft and different firms. The Hawker case studies are for the same company and the majority of case studies show real cost escalation, with escalation factors varying between 1.07 and 3.20. Next, questions arise as to whether cost escalation is confined to defence projects.

11.1 Examples

A 1977 study estimated cost escalation and its determinants for a variety of sectors, including military aerospace, building and civil engineering and the private sector. Cost escalation was found in all these sectors (Hartley and Cubitt, Reference Hartley and Cubitt1976–77). Nor was it new; it occurred on nineteenth-century projects such as canals, railways, tunnels and bridges. Within the private sector, it was found in the chemicals, computer, electronics, engineering, pharmaceutical and textile industries. A study of cost escalation in the Norwegian road construction industry between 2009 and 2014 found cost escalation averaging 17 per cent. The main cause of escalation in road construction was change orders to the scope that were not covered by the original contract. There are other examples from the Olympic Games, the Panama Canal and the ITER (International Thermonuclear Experimental Reactor) project. The 2012 London Olympic Games were initially estimated to cost £2.4 billion, but the eventual cost was £8.8 billion. The Panama Canal extension had an initial cost estimate of $5.25 billion for completion by 2014, but its final cost was $6.1 billion in 2016. The ITER project to demonstrate nuclear fusion as a source of clean energy, which involved thirty-five countries, had an initial cost estimate of €5 billion for completion in 2016, but the latest estimate was a cost of €20 billion to be completed by 2025.

A good example of the procurement problems for UK construction projects is HS2. The project was for a high-speed railway line running from London to Manchester and Leeds. In 2011, it was estimated to cost £32 billion; by 2013 its costs had risen to £55.7 billion; and by 2019 the estimated costs were almost £80 billion with completion expected by 2033. Against this background, the northern section from Birmingham to Manchester was cancelled in October 2023 after long delays and cost overruns as well as estimates that it would cost over £70 billion in 2019 prices, with some estimates suggesting costs as high as £170 billion. Also, following the pandemic, the business travel benefits of the project had changed and the project no longer provided ‘good value for money’.

11.2 UK Government Advice

The UK government, through the Infrastructure and Projects Authority (IPA), published Cost Estimating Guidance in 2021 (IPA, 2021). It stressed the use of evidence-based cost estimates which are not a single figure but one that evolves over time as a range, depending on the level of risk and uncertainty. It also stressed that at the heart of a robust cost estimate are the three Ps, namely, principles of cost estimating, people involved in the estimate and performance reflected in the quality and consistency of cost estimates. The document outlined a best practice approach to cost estimating. One result of cost uncertainty for contractors was the inclusion of a price escalation clause in contracts. In the UK construction industry in 2021–22, the price indices for ‘All Work’ increased by 27.2 per cent to May 2022, with the largest increase of 64.9 per cent for concrete reinforcing bars (Weaver, Reference Weaver2022; Welde and Dahl, Reference Welde and Dahl2021).

Estimates of cost escalation for civil projects have some limitations. They might be based on current rather than constant prices and sometimes the price basis is not clear. Figures might be compared at different stages in project development (e.g. during initial design work and when nearing completion) and private sector projects might be shrouded in secrecy (e.g. private firms might be unwilling to reveal details of secret or cancelled projects).

12.1 The Predictions

Aircraft unit cost curves are expected to be downward sloping, showing falling average costs with increased output. Costs are of various types: short and long run; fixed and variable; average and marginal; and money and real (opportunity costs). The costs in this study are short and long run, fixed and variable, average and marginal and money expenditures. The contracts differ between aircraft and they comprise fixed (overheads) and variable costs (labour, materials, flight testing); some contain the costs of jigs and tools, different quantities ordered, and in some cases sufficient information is provided to enable estimation of marginal costs.

For some aircraft, there were second sources of supply so that unit cost data were available for alternative suppliers. These were World War II aircraft and those produced in the immediate post-1945 period. These data allowed an evaluation of the efficiency of procurement policy.

12.2 Learning Economies

Learning economies are a further source of decreasing cost for aircraft manufacture. Repetition leads to decreasing costs, with workers learning from experience and productivity rising with output. This effect is shown by data on unit prices, unit production costs and unit labour costs. Learning affects all these variables. Unit labour costs show the direct effects of learning, although these were not always available, hence the use of unit price and unit production cost data. The published data on prices, unit production costs and labour costs also have their limitations, which have to be borne in mind when interpreting the results. Published data are not necessarily least cost or X-efficient. They are simply observations and are not necessarily points on the same efficient average cost curve. Instead, they could be points on efficient cost curves, or points on inefficient curves or points on different curves reflecting different efficiency levels. These limitations apply to the empirical work reported in this study.

12.3 The Evidence

Evidence on unit cost curves is presented below for a variety of war and peacetime aircraft. The data are presented in two sections, namely, in the sections that follow and in the Statistical Appendices. The data presented here are for major wartime aircraft, namely, the Spitfire and the Lancaster, followed by post-1945 aircraft, the Meteor, the Hunter, the Lightning and two bombers, the Canberra and the Valiant. The data in the Statistical Appendices show supporting information for pre-war aircraft such as the Hind and the Hampden, for World War II aircraft, namely, the Hurricane, the Halifax and the Mosquito and for post-1945 aircraft such as the Vampire, the Venom and the Javelin and two V-bombers. The Spitfire is the first aircraft in the analysis.

12.4 Spitfire Fighter

The Spitfire was the classic World War II fighter aircraft which, with the Hurricane, fought in the Battle of Britain (details of the Hurricane appear in the Statistical Appendices). The Spitfire was a battle-winning aircraft with its contributions in the Battle of Britain and the rest of World War II (i.e. a successful aircraft). Cost-quantity data for the Spitfire are shown in Table 7. Spitfire unit prices and costs generally declined with greater output, although there was some evidence of ‘tailing-off’ towards the end of production. Westland, as the second supplier, was a costlier new entrant compared with Supermarine: for example, unit labour costs were £700 for Supermarine compared with £1,500 for Westland for early production contracts (in constant prices). The cost differentials indicate the price that the UK was willing to pay for a second source of supply. Profit margins for the original developer, namely, Supermarine, were also higher than Westland in early production.

12.5 Vickers Wellington Bomber

The Wellington bomber was included in the analysis because it achieved the largest output of all UK bombers: over 11,400 aircraft were produced in World War II. Cost details are shown in Table 8.

The Wellington was a decreasing cost aircraft, especially for unit production and unit labour costs: unit labour costs for the final output were some 60 per cent of those for the initial output. Profit rates also declined with output, falling from 7.5 per cent for the early output to 3.5 per cent for the final contracts (which was typical of wartime contracts).

12.6 Avro Lancaster Bomber

The Lancaster bomber is included because it was one of the most famous aircraft of World War II. Its cost details are shown in Table 9.

Unit costs for the Lancaster declined continually with greater output, except for a ‘tailing-off’ effect approaching the end of Lancaster production (see unit production costs, Table 9). The costs of new entry were substantial, as shown by initial price and cost data for Avro and Armstrong Whitworth: for example, for early production, Armstrong’s unit prices were some 65 per cent higher than Avro’s. Profit rates for Armstrong Whitworth also varied, rising towards final output.

12.7 Gloster Meteor Jet Fighter

Data for the Meteor are presented because it was the UK’s first jet fighter, followed by the Hunter and the Lightning. Meteor unit cost and price curves are shown in Table 10.

For the Meteor day fighter there were substantial reductions in prices and costs over total production: the 1952 prices and labour costs were 32 per cent of those in 1946 and unit production cost reductions were even greater. The night fighter version showed reductions for unit prices but increases for unit production and labour costs.

12.8 Hawker Hunter

The Hunter is included as a first-generation UK supersonic fighter. Its unit cost curves can be compared and contrasted with those of the Meteor (Table 11).

The unit price data show Hawker unit costs declining initially and then rising, followed by either a further rise or a fall, depending on the relevant cost curve. Overall, between 1955 and 1957, unit price and production costs declined to some 85–84 per cent and labour costs fell to 71 per cent of their initial levels. Compared with the Meteor, the reductions in prices and costs for the Hunter were much less. This probably reflected smaller cumulative output and less learning for the Hunter.

12.9 English Electric Lightning

The Lightning is included as a third-generation fighter capable of much greater speeds than the Meteor and the Hunter. The unit price and cost curves for the Lightning declined initially, then generally rose, with some higher costs reflecting a shift to trainer aircraft and the reorganisation of the UK aircraft industry (Table 12). Certainly, after the second contract there is no evidence of decreasing costs. Profitability varied between contracts. Interestingly, it reached its highest level for the final contract: this is surprising since it was expected that profit rates would decline towards the end of production.

12.10 Jet Bombers

This section focuses on jet bombers and covers the Canberra and the Valiant. The Statistical Appendices cover the other V-bombers, namely, the Victor and the Vulcan.

12.11 The Efficiency of Procurement

Overall, the evidence presented shows aircraft as a decreasing cost industry. Cost reductions were especially substantial for unit labour costs, which are more accurate reflections of learning curves.

The evidence on second suppliers allows an assessment of the efficiency of alternative suppliers. Typically, alternative suppliers were higher unit cost firms and the additional costs were substantial. For example, on the Spitfire and the Lancaster, unit labour costs for the second supplier were about double the costs for the original developer. But this was the price the UK was willing to pay for a second supplier in wartime, where the emphasis was on availability and supply rather than a narrow interpretation of efficiency.

A similar problem occurred with the efficiency of the procurement of the UK’s three V-bombers. Efficiency, interpreted narrowly, required the procurement of one type of V-bomber rather than three types. A total of 435 V-bombers were purchased in their three variants. If 435 units of one type had been purchased there would have been savings of two R&D bills and the economies of scale and learning from a longer production run. To illustrate the possibilities, assume that only the Victor had been purchased. The savings that would have been achieved through not needing two extra R&D bills would have totalled some £10 million and unit production costs would have been reduced by 20 per cent or some £23.3 million (at 1955–56 prices). The total savings in R&D and productions costs by purchasing just one type of V-bomber might therefore have been around £33 million, which was a substantial number of sacrificed schools, hospitals and roads (at 1955–56 prices). This was the price paid for an insurance policy provided by selecting three V-bombers. Economic efficiency requires that these extra costs be viewed as representing a worthwhile investment.

13.1 The Sample

Unit price and cost escalation data for non-combat aircraft, namely, trainer aircraft are shown in Tables 16 and 17. Unit prices have risen from some £1,400 for the 1939 Tiger Moth to £233,400 for the supersonic Lightning in 1960, a rise by a factor of 168.0 in real terms over the whole period.

13.2 Cost Escalation on Trainers

Typically, cost escalation is associated with advanced combat aircraft, both fighters and bombers. Questions arise as to whether cost escalation arises with other groups of aircraft such as various types of trainer aircraft comprising basic, advanced and multi-engine types. Cost escalation factors are estimated for a group of propeller-powered basic trainer aircraft, except for the Jet Provost, which is included in the group as it was also a basic trainer (reflecting a shift to all jet trainers for the RAF). It can be seen that basic trainers showed both real cost increases and decreases (Table 17). Over the period 1938 to 1957 (almost twenty years), the real unit costs of basic trainer aircraft increased by a factor of slightly over 4.0. Cost escalation factors were considerably higher for jet-powered advanced trainer aircraft, reaching a factor of almost 11.0 over the thirteen-year period from 1947 to 1960 (see Table 17). Similarly, cost escalation was also estimated for multi-engine trainers between 1938 and 1952. Using a small sample, real cost escalation factors of between 2.3 and 3.1 were estimated for larger trainer aircraft. Overall, this section confirms that trainer aircraft were also subject to real cost escalation and that escalation is not confined to advanced combat aircraft.

14.1 The Hypotheses

This section provides a further opportunity to test some of the hypotheses suggested by Augustine. Various rank correlations and regressions were estimated to test for Augustine weapons effects. Tests were undertaken of the determinants of unit prices for various types of aircraft.

First, for fighters and bomber aircraft, rank correlations were estimated between unit prices and each of weight, speed and time. Unit prices represented unit costs since military aircraft prices are usually estimated on the basis of unit production costs plus a profit margin. All price and unit costs were in constant prices (1959). Weight was all-up weight or the gross weight of an aircraft, measured in pounds (lb). Speed was aircraft speed in miles per hour (mph). For fighters and bomber aircraft, Spearman’s rank correlation between unit prices and weight was +0.78, between unit prices and speed was +0.58 and between unit prices and time was +0.73, each showing positive and significant relationships between the variables. For fighters and bombers, greater weight and speed led to higher unit costs and higher unit prices, with real unit prices rising over time.

14.2 Regressions

Regressions were estimated for three different groups of aircraft, namely, fighters, fighters and bombers, and all combat aircraft, including large aircraft (Samples 1–3: time-series estimates). The results are summarised in Tables 18 and 19. Equation (1) in Table 18 (Sample 1) estimates unit prices for fighter aircraft against weight, speed and a linear time trend. The equation accounts for some 93 per cent of the variability in unit prices with significant coefficients for weight, speed and time (Equation (1), Table 18). However, the coefficient on time had an unexpected negative sign. When the equation was re-estimated (Equation 2, Table 18) without the speed variable, it still resulted in a negative sign for time but this was not significant. Further estimates found speed significant, with a positive effect on unit prices for fighter aircraft (Equation (3), Table 18).

Different results for time emerged when the sample was widened to include both fighter and bomber aircraft (Sample 2). In the wider sample, both weight and time had significant and positive impacts on unit prices with the equations explaining over 70 per cent of the variations in unit prices (Table 19, Equations (4)–(7)). Elsewhere in Sample 2, the variables for speed and a dummy distinguishing between fighter and bomber aircraft were not statistically significant (Table 19, Equations (5) and (6)).

Overall, the empirical results found that aircraft weight was a major determinant of unit prices and there was some evidence of unit prices rising with time. Augustine found different results. He found prices rising with time and no impact on prices of aircraft characteristics such as weight and speed.

14.3 A Larger Sample

Finally, the sample was widened further to include large aircraft: Table 20 (n=38). The explanatory value of the equations with the larger sample was reduced substantially, explaining only some 30 per cent of price variations. Also, time was the only significant coefficient with the predicted positive coefficient: neither weight nor speed was significant, with the results confirming Augustine’s original findings. Also, the Pearson correlation coefficient between unit prices and time was positive and significant,Footnote ⁴ confirming one of Augustine’s predictions.

The empirical results presented in this section are exploratory, testing some of Augustine’s hypotheses using time-series equations. Contrary to Augustine’s findings of time as a significant factor, there was evidence that aircraft characteristics in the form of weight and speed were also statistically significant. Augustine claimed that unit costs were closely correlated with time rather than with changes in speed, weight or other technical parameters. Undoubtedly, there remains scope for much further work in this field.

15.1 Characteristics of Augustine Weapons

Augustine weapons systems are characterised by the following features:

1. 1. For advanced combat aircraft, rising unit costs correlate closely with the passage of time rather than with changes in speed, weight or other technical performance characteristics. According to Augustine, the trend of rising unit costs also applies to commercial aircraft, helicopters, ships, tanks, automobiles and homes, but at a lower growth rate: a factor of 2.0 every ten years.

2. 2. Cost and budget trends led Augustine to forecast that by 2054 the entire US defence budget will purchase just one aircraft. Similarly, the UK will reach the single-aircraft position almost two years before the USA.

3. 3. Rising costs reflect the use of complex electronics, computer software and the growing application of stealth technologies which open up vast new capability vistas which are then crammed into each new generation of a product.

4. 4. The engineering mindset of decision-makers in the military industrial complex values technology for technology’s sake, regardless of its cost effectiveness. Critics claim that major weapons systems are laden with technological ‘bells and whistles’ which add considerably to costs but little to military effectiveness.

15.2 Correlations and Causation

A starting point in critiquing Augustine is that much of his analysis is based on extrapolating historical cost data, which presents the possibility of confusing observed trends with causation. Correlations are not evidence of causation. The problem with trends is that they bend: there is a famous saying that a trend is a trend but the question is will it bend and alter its course through some unforeseen force and come to a premature end! The possible end cannot be ignored for Augustine systems.

Augustine cost extrapolations have been used by critics of the military industrial complex as proof of wasteful military spending, gold plating and defence establishments with little incentive to control costs. They claim that military investment is captured by complex vested interests within the MIPC and that much new defence capability is over-engineered and wasteful, leading to smaller quantities of increasingly complex and costlier equipment. Nor do we focus on the benefits side of military investment, examining whether these ever costlier Augustine investments lead to more defence being produced by each incremental dollar. Here, there is a major deficiency: there is an absence of a money value of defence output. Instead, typically defence output is measured by inputs and the assumption that inputs equal outputs. On this basis all increases in defence inputs mean higher defence outputs, which is a far from satisfactory position!

15.3 Lack of a Theory

There remains a major deficiency, namely, that the Augustine approach lacks a well-established theory in the form of a set of definitions and assumptions which provide logical deductions explaining behaviour. Instead, there are assertions and anecdotal evidence which provide a good story but do not represent a well-established theory. Augustine’s ‘model’ comprises a set of observations which have been joined together to form a narrative which is appealing and attractive but by itself does not form a theory.

16.1 Balance between Augustine Weapons and Cheap Drones

What are the implications of Augustine weapons for the future of defence industries and the armed forces? The high-technology feature of Augustine weapons will lead to a smaller number of larger firms which will be R&D-intensive rather than production-intensive. The search for markets and production orders probably means more international mergers. But these trends will be offset by the emergence of small and cheap drones that are capable of a variety of military tasks. For example, drones might be used for surveillance and observation, thus replacing the military personnel traditionally used for such tasks. They also provide a long-range strike capability, so enabling remote warfare (e.g. the ability to hit targets without military occupation of the country). But with cheap drones there will be pressures to make them more complex so as to perform ever-more-complex tasks at ever-increasing costs: ultimately, they will become as costly as Augustine weapons.

So, where is the likely balance between expensive Augustine weapons and cheap drones? Currently, there is a role for both. Substitution effects might lead to cheap drones replacing costly Augustine weapons. Alternatively, drones and Augustine weapons might be complementary, with costly Augustine weapons controlling a number of drones in a range of strike and reconnaissance missions. The model here is swarms of cheap drones commanded by costly Augustine weapons (the USA created Unit X as part of the Pentagon to establish links with Silicon Valley and transform the future of war: Shah and Kirchhoff, Reference Shah and Kirchhoff2024).

Both Augustine weapons and drones have impacts on the armed forces. Costly high-technology Augustine weapons mean smaller numbers, with capital and technology replacing military personnel. Already, some nations are finding it too costly to replace their combat aircraft fleets and have chosen to opt out of maintaining combat air capabilities (e.g. New Zealand). There are implications for the training of military personnel, with probably a greater reliance on the use of simulators for training. High costs mean greater incentives for national armed forces to share such weapons, creating international military forces and capabilities. The emergence of cheap drones creates opposite trends to the Augustine effects. Like Augustine weapons, they are technology-intensive, but their relative cheapness means that they can be purchased in large numbers. There will be substitution effects for the armed forces as drones replace some of the traditional roles undertaken by manned systems (e.g. strike and reconnaissance missions). Training requirements in the armed forces will change as military personnel use computers to acquire skills compared with traditional battlefield experience. Skilled labour will be in greater demand as military personnel will need to be more skilled to cope with complex equipment. Future demands for military personnel will be for highly skilled inputs and there will be concerns about the armed forces obtaining a worthwhile return on their costly training investments (Arce, Reference Arce2023).

16.2 International Collaboration

The continually rising costs of Augustine weapons create greater pressures for international collaboration to share costly development and combine small production orders. The costs of Augustine weapons means that they are tailor-made for international collaboration. For example, currently European nations are developing two types of future combat aircraft, namely, the UK-led Tempest forming the Global Combat Aircraft Programme, comprising the UK, Italy and Japan, with an in-service date of 2035 and the Franco-German-Spanish sixth-generation fighter aircraft.Footnote ⁵ Questions arise as to whether these nations can afford to develop two similar independent projects to be bought in small numbers. Whilst international collaboration is economically attractive, it has political costs: who will be the project leader and how will work be shared between the partner nations?

International collaboration benefits both development and production work. Development work is a fixed cost which can be shared between two or more nations. For example, consider a project with development costs of £20 billion: without collaboration a single nation would pay the fixed cost of £20 billion, but shared among four nations the bill for each nation is only £5 billion (ceteris paribus). Similarly, sharing production work has benefits. Each nation might demand, say, 100 aircraft; a four-nation collaboration would therefore mean a production order for 400 aircraft rather than the 100 aircraft required by one nation, and an order for 400 aircraft results in lower unit costs from economies of scale and learning.

But this simple economics of collaboration suffers from one flaw: the ceteris paribus assumption. With international collaboration, other things do not remain unchanged and the assumption of fixed development costs remaining unchanged at £20 billion is most likely to be null and void. Nations will have their national requirements for an advanced technology project and each will demand to be involved in the advanced technology part of the programme (e.g. fuselage, engine, avionics) and to get their ‘fair share’ of the work. As a result, the work share or ‘fair shares’ requirement is likely to depart from efficiency criteria. For example, each nation will demand a flight test centre and a final assembly line and similar duplication will arise in the management of the collaboration. An official UK report on the collaborative Typhoon programme concluded that collaborative decision-making was inefficient (NAO, 2011). These are not costless additions to the collaboration and will increase total development costs beyond the original £20 billion used in this example. But international collaboration will continue to be economically worthwhile so long as its costs are less than the costs of an independent national development (in this example, £20 billion). Or collaboration is worthwhile so long as it costs less than the costs of importing from the least cost foreign supplier. In these circumstances, cost–benefit analysis can be applied. But judgement cannot be avoided, especially in valuing the costs and benefits of the intangible elements such as national independence and the timescale of the assessment. Cost estimates will vary and be subject to different degrees of reliability. For example, cost estimates at the start of a project will be less reliable than later estimates, though often initial cost estimates are used as the basis for a project decision.

The Airbus A400M airlifter is a good example of an international collaboration involving a large number of nations. Initially, it was expected that Airbus as a successful civil airliner firm would have few problems in developing a military airlifter. Reality was different. The airlifter was a seven-nation development using a new engine and composite technologies. Since its launch in 2003 it has been characterised by cost overruns and considerable delays. Costs have risen by about €8 billion on a project originally expected to cost €20 billion and the first aircraft was delivered four years late. It has also failed to meet its capability requirements, leading to further penalties. The original contract assumed wrongly that commercial terms could be applied to a complex international military aircraft procurement. Problems remain with the engine and the gearbox and Airbus admits that it under-estimated the complexity of the A400M.

For the future, the next development in technology might well be the development of space forces with the capability of destroying communications and surveillance satellites. Such satellites provide information to Earth-based armed forces, so their destruction would provide potential enemies with a battle-winning advantage (Markowski et al., Reference Markowski, Brauer and Hartley2023).

16.3 Role of Humans

There is a further factor which cannot be ignored, namely, the role of human beings in Augustine weapons. Despite their massive complexity, such weapons are still operated and commanded by humans with all their limitations: humans, unlike popes, are not infallible. Artificial intelligence (AI) means the emergence of different types of military commander – AI commanders – having available vast amounts of data for their decision-making. Machines will support human decisions and might even replace them by making decisions. An example illustrates the possibilities. In 1940, Fighter Command managed the Battle of Britain with radar, ground spotters and command centres which alerted aircraft to the threats from the attacking force. Today, the Israeli Iron Dome air defence system manages air defence without human intervention: the trajectory of incoming rockets is tracked and only those likely to reach important targets are selected for interception. The system is currently expensive; the costs of each interception are high, although the attacking drones and rockets themselves are relatively inexpensive (Freedman, Reference Freedman2023). Using AI means that decisions will not be dominated by inter-service rivalries and personalities (e.g. Montgomery and Patton in World War II); however, such disputes and differences in judgement are sure to occur in different forms.

16.4 Communications

Military communications also remain important. The ability to communicate quickly and accurately with military staff in different locations will continue to be important even with Augustine weapons. And communications have been subject to technical change, varying from the historical examples of communications depending on human runners taking messages between command centres to flag signals between ships and the use of pigeons! Communication failures have determined the outcomes of battles (e.g. Arnhem). Modern military communications are much more complex, involving computers, information technology, networks and space-based satellites, and these systems need to be protected (cyber security). Allocating Augustine weapons in battlefield situations will depend on communications. Costly Augustine weapons can be destroyed by faulty communications; the systems are so costly that they might require even costlier systems for their protection!

16.5 The Future

What of the future for Augustine weapons? Futures are characterised by uncertainty: today’s winners are tomorrow’s losers. New technology has dominated military weapons and some technical change has been revolutionary as well as evolutionary. The jet engine is an example of revolutionary technical change rendering obsolescent many previous engine technologies. Most defence equipment is made to order; thus, for military firms to develop and produce the capabilities needed for Augustine weapons, the risks will be largely shifted to taxpayers. This creates new regulatory problems, which will be international as firms increasingly seek international mergers, and there will be new opportunities for rent seeking. All of this suggests that taxpayers will be at even greater risk of exploitation by the MIPC.

Augustine weapons are costly and will become costlier. But they are procured against a background where wars are also costly. The costs of war are direct and indirect. Direct costs include deaths and injury amongst military and civilian personnel for all the participants. There are also costs to physical structures in the destruction and damage to buildings, homes and infrastructure (e.g. roads, railways and bridges). Air strikes will affect populations in terms of their location, movement and shelter. The costs become greater where a population is invaded and subdued. Invasion means a loss of freedoms and liberty, resulting in starvation and slavery. The Nazi German occupation of Europe and the Soviet Union in World War II is a good example. Once the costs of war are considered, expensive Augustine weapons might be regarded as worthwhile: they are cheap at the price.

Augustine weapons are not unique. The term applies to current weapons, but there are historical examples. Castles are the medieval equivalent. They were costly and at the time were superior, battle-winning weapons. But new technology in the form of cannons rendered them obsolescent; and in the military field, new technologies are always emerging. Previous generations of battleships were the Augustine weapons of their day, but along came aircraft and aircraft carriers, which rendered them obsolescent with nuclear weapons having similar effects. Regardless of the age, all weapons have one distinctive characteristic: they are designed to kill people. Augustine weapons have a special added characteristic in that they are much costlier than their predecessors.