Why innovation outcomes differ among defence innovation systems: a comparative study of radar innovation in Sweden and Australia
Robert Charles Wylie
A thesis in fulfilment of the requirements for the degree of Doctor of Philosophy
School of Business
UNSW@Canberra
October 2014
ABSTRACT Why do nations at comparable stages of economic development, with comparable political systems and with access to comparable technologies perform differently in generating novel solutions to similar requirements for military capability? To address this question the thesis compared case studies of radar-based innovation in Sweden and Australia during the Cold War. The case studies were organised around the "building blocks" of a defence sectoral system of innovation which comprised institutions, actors and networks, military doctrine, technology and the exercise of demand. Development of innovative surveillance radars in, respectively, Sweden and Australia was then used to show how the functioning of those building blocks influenced the performance of the Swedish and Australian innovation systems. The performance of each system was then compared in terms of the time each took to develop their respective radars, the cost they incurred in doing so and the development/diffusion of those radars after their acceptance into Swedish and Australian service respectively. The comparison showed that distinctive features of each country's defence sectoral innovation system caused Australia to take longer than Sweden to develop a broad area surveillance radar, to incur higher costs in doing so, to pursue a narrower path of post-acceptance development of the radar and to impose more stringent constraints on the diffusion of the resulting technology. The thesis makes a novel contribution to the literature on, and to the management of, military technological innovation in terms of the subject addressed, the methodology used and the conclusions reached.
Table of Contents Page
Acknowledgements v
Abbreviations and Symbols vi
List of Figures ix
List of Tables x
Chapter 1 Introduction 1
1.1 Thesis Limits: Military innovation in small democracies. 2
1.2 Thesis Focus: Military capability and military technological innovation. 3
1.3 Thesis technology focus: Radar. 4
1.4 Thesis Metrics: Innovation performance. 4
1.5 Thesis Framing: A system perspective. 5
1.6 Thesis Research: Case Study Methodology. 6
1.7 Thesis Structure. 9
Chapter 2 Literature 11
2.1 Defining innovation and military technological innovation. 11
2.2 The defence innovation literature 12
2.3 The wider innovation literature. 18
2.4 The innovation systems literature 21
2.5 Conclusion 29
Chapter 3 Defence Sectoral Innovation Systems 30
3.1 Defining and ordering the defence innovation building blocks 30
3.2 Identifying the distinctive features of defence innovation building blocks 34
3.2.1 Institutions. 34
3.2.2 Actors and Networks. 36
3.2.3 Defence-specific knowledge 42
i
3.2.4 Technology. 43
3.2.5 Demand. 51
3.3 Conclusion 53
Chapter 4 The Swedish defence sectoral innovation system 56
4.1 Swedish institutions 56
4.2 Swedish actors, networks and the Swedish competence block. 62
4.3 Swedish military doctrine. 72
4.4 Swedish technology. 74
4.5 Swedish demand. 83
4.6 Conclusion. 86
Chapter 5 Developing, Procuring, Operating and Diffusing the ERIEYE Radar System. 87
5.1 Developing ERIEYE: Overview. 87
5.2 Developing radar competencies at Ericsson 88
5.3 Developing a rapid reaction surveillance radar. 92
5.4 The requirement for a rapid reaction surveillance system 97
5.5 Executing demand for a rapid reaction surveillance system. 98
5.6 Post-acceptance ERIEYE diffusion. 105
5.7 Conclusion. 107
Chapter 6 The Swedish innovation system and ERIEYE innovation outcomes
108
6.1 The influence of Swedish Institutions 108
6.2 The influence of the Swedish defence competence block. 111
6.3 The influence of Swedish military doctrine. 116
6.4 The influence of the Swedish technology base. 119
6.5 The influence of Swedish demand. 122
ii
6.6 Conclusion 125
Chapter 7 The Australian defence sectoral innovation system 126
7.1 Australian institutions. 126
7.2 Australian actors, networks and the Australian competence block. 133
7.3 Australian military doctrine. 143
7.4 Australian technology. 146
7.5 Australian demand. 151
7.6 Conclusion. 154
Chapter 8 The Jindalee Operational Radar Network 155
8.1 Developing JORN: Overview. 155
8.2 Initial investigation of OTHR technology. 157
8.3 Project Jindalee: Developing OTHR for broad area surveillance. 160
8.4 Formulating demand for an OTHR-based broad area surveillance capability. 165
8.5 Executing demand for an OTHR-based broad area surveillance capability. 168
8.6 Post-acceptance development of JORN. 181
8.7 Conclusion. 185
Chapter 9 The Australian innovation system and JORN innovation outcomes
187
9.1 The influence of Australian institutions. 187
9.2 The influence of the Australian defence competence block. 190
9.3 The influence of Australian military doctrine. 201
9.4 The influence of the Australian technology base. 204
9.5 The influence of Australian demand. 206
9.6 Conclusion. 213
iii
Chapter 10 Explaining divergent innovation performance. 215
10.1 Military technological innovation in Sweden and Australia 216
10.1.1 Comparing Swedish and Australian institutions. 10.1.2 Comparing Swedish and Australian competence blocs. 219
10.1.3 Comparing Swedish and Australian military doctrine. 223
10.1.4 Comparing Swedish and Australian technology. 224
10.1.5 Comparing Swedish and Australian demand. 227
10.2 Contribution to military technological innovation literature. 229
10.3 Managing military technological innovation 232
10.3.1 Undertaking military technological innovation. 232
10.3.2 Learning for military technological innovation. 233
10.3.3 Exchanging information for military technological innovation. 234
10.3.4 Allocating risk and accountability for military technological Innovation 234
10.4 Conclusion and directions for future research. 235
Bibliography 237
Interviews 248
Parliamentary Hearings 250
Thesis-related publications and presentations 251
iv
Acknowledgements This thesis would not have been started without the inspiration of Sir Malcolm Kenneth McIntosh (1945-2000). It would not have been completed without the support of my interviewees who shared their stories, my supervisors who guided my work, my colleagues who shared my journey and my family who encouraged and sustained me.
The thesis taps the experience of innovators in Sweden and Australia. Among the many Swedes who gave me their time, shared their knowledge and sponsored me into their networks my deepest thanks go to Sven Larsson, Carl-Gilbert Lonroth, Lennart Kallquist and Charles Edquist. Of the numerous Australians who shared their knowledge and insights, alerted me to nuance and pointed out my errors, I must single out Don Sinnott and Paul Johnson for special thanks.
The thesis reflects the guidance and support of my academic mentors. My heartfelt gratitude to Professor Peter Hall who supervised my work with patient insight, guided my learning over innumerable cups of coffee and helped me clarify my thinking with unfailing grace. Thanks also to Professor Michael O’Donnell who somehow always made time to read and re-read my drafts and to Associate Professor Stefan Markowski who never failed to stimulate me with comments and ideas.
The thesis would never have been finished without the practical support of the willing professionals who work in the UNSW@Canberra library, in the Committee Secretariat for the Australian Senate, in Australia’s National Library and in Sweden’s Riksdag Library. I also acknowledge the cheerful encouragement of my colleagues in the School of Business, UNSW@Canberra. By demonstrating that the job could be done they helped me do it.
My thesis was both personal journey and academic endeavour. I cannot adequately express my gratitude to Sylvie, my wife, who was always there as I struggled to make sense of my professional experience. Nor can I overstate my thanks to our two children, Neil and Nicole. As recently qualified medical doctors they understood - perhaps better than I did - what the endeavour entailed.
v
Abbreviations and Symbols AEW&C Airborne Early Warning and Control
AII Australian Industry Involvement
ANAO Australian National Audit Office
AN-APG 55 Pulse Doppler intercept radar fitted to the RAAF F/A-18 aircraft
AN-FPS 118 Over-the-horizon backscatter radar (originally developed for the US Air Force)
AN-TPS 71 ROTHR Relocatable over-the-horizon backscatter radar (originally developed for the US Navy)
AOCI Australian Ownership, Control and Influence
ARPA Advanced Research Projects Agency (US)
ASCC Air Standardisation Coordinating Committee
ASEAN Association of South East Asian Nations
ASIC Air and Space Interoperability Council
AWA Amalgamated Wireless Australasia (company)
BAE British Aerospace (company)
BHP Broken Hill Proprietary Limited (company)
CAFOP Chief of Air Force Operations and Plans
COCOM Coordinating Committee
CSA Computer Sciences Australia
CSIR Council for Scientific and Industrial Research
CSIRO Commonwealth Scientific and Industrial Research Organisation
CTD Capability and Technology Demonstrator (a fund administered by the Defence Science and Technology Organisation)
DAO Defence Acquisition Organisation
DDAW Defence Designated and Assisted Work
vi
DEPSECA Deputy Secretary (Acquisition)
DID Defence Industry Development (a fund administered by the Australian Department of Defence).
DSDC Defence Source Definition Committee
DSTO Defence Science and Technology Organisation
Draken a second-generation jet fighter designed and built by SAAB with integrated weapon and avionics systems and a distinctive double delta wing configuration (Full designation SAAB J35 Draken)
FFA Flygtekniska försöksanstalten (Swedish Institute for Aeronautic Research)
FFVS Flygförvaltningens Verkstad (Aeronautical Studies Workshop)
FMV Försvarets materielverk (Swedish Materiel Agency)
FOA Försvarets forskningsanstalt (Swedish Defence Research Establishment)
FOI Totalförsvarets forskningsinstitut (Swedish Defence Research Institute)
IP intellectual property
JFAS Jindalee Facility Alice Springs
JPO JORN Project Office
JSET Joint Service Evaluation Trials
Gripen , a single-engine multirole fighter aircraft manufactured by SAAB (Full designation SAAB JAS 39 Gripen)
Lansen a single engine all weather jet fighter aircraft manufactured by SAAB in various configurations (Full designation SAAB J32 Lansen)
NRL Naval Research Laboratories (US)
MADRE Magnetic Drum Radar Equipment (developed by NRL)
MIF Melbourne Integration Facility (built by RLM for JORN)
MUL Multi-user list (of contractors pre-approved for Australian defence procurement)
NATO North Atlantic Treaty Organisation
vii
RAF Royal Air Force
RAP Recognised Air Picture
OTH-B Over-the-horizon – Backscatter (radar)
OTHR Over-the-horizon radar
Riksdag The Swedish national legislature
SAAB Svenska Aeroplanactiebolaget
STRIL A Swedish command, control, communication system
STRIC A Swedish command, control, communication system
SwAF Swedish Air Force (Flygvapnet)
TTCP The Technical Cooperation Program
Tunnan The Swedish Air Force’s second jet engine powered fighter, built by SAAB
viii
List of Figures Page
Figure 3.1 ‘Value for money’ in British usage 52
Figure 5.1 ERIEYE development overview 88
Figure 5.2 Ericsson/SAAB: selected airborne radar development and production 91
Figure 5.3 ERIEYE coverage for airborne early warning 104
Figure 8.1 Overview of JORN development, procurement and upgrade 156
ix
List of Tables Page Table 5.1: Developing Ericsson competency in pulse Doppler radar technology 94
Table 5.2: Developing Ericsson competency in electronically scanned array technology 97
Table 5.3: Developing and producing Erieye for the SwAF 105
Table 5.4: Adapting ERIEYE for export 107
Table 7.1 Australian defence industrialists 1986 (by defence industry sector) 142
Table 8.1: Opportunistic research 159
Table 8.2: Project Geebung 160
Table 8.3: Jindalee Stage A 162
Table 8.4: Jindalee Stage B 164
Table 8.5: Jindalee Facility Alice Springs (JFAS) 165
Table 8.6: From JFAS to JORN 168
Table 8.7: Executing demand for JORN 181
Table 8.8: JORN post acceptance integration and upgrade 185
Table 10.1 Comparing JORN and ERIEYE innovation outcomes 215
x
Chapter 1: Introduction
This thesis investigates the following question: Why do nations at comparable stages of economic development, with comparable political systems and with access to comparable technologies perform differently in generating novel solutions to similar requirements for military capability?
This research question is particularly topical in Australia. Past controversies over Australian military technological innovation colour current Australia debates about, for example, the appropriate mix of indigenous ‘make’ and overseas ‘buy’ solutions to Australian requirements for military capability. Hence the thesis seeks to assist practitioners involved in these current policy debates by suggesting a framework that might help them bring the issues involved into coherent focus.
The research question is also of interest to scholars of military technological innovation, much of whose work has proceeded, as James has observed, with only weak links to the concepts and theories recently developed in the wider innovation literature.1 Hence the thesis aims to contribute to the military technological innovation literature by demonstrating the analytical utility of certain concepts developed and refined in the wider innovation literature relatively recently. Equally, however, the research question is of interest to those scholars engaged in developing concepts and theories to describe and explain the wider innovation phenomenon, much of whose work makes at best passing reference to military technological innovation. Hence the thesis also seeks to contribute to the wider innovation literature by demonstrating how certain generic innovation concepts can be adapted so as to shed additional light on the role of design-based learning in the wider innovation process.
As a first step in interpreting the above research question, this chapter begins by limiting the thesis to a particular category of nations to which the question is addressed. Two such nations are identified and the reasons for their selection explained briefly. The second section of the chapter explains the references in the research question to the notions of ‘military capability’, and to ‘requirements for military capability’. The third section of the chapter addresses that element of the research question concerning access to comparable technologies. The thesis is about comparing innovation performance based on a generally available technology. This thesis uses radar as a generally available technology to facilitate investigation of why similar nations addressing similar requirements for military capability in similar timeframes can generate divergent innovation outcomes. Accordingly, the third
1 Andrew D. James, Re-evaluating the role of military research in innovation systems, Journal of Technology Transfer, Vol 34, 2009, especially, pp. 451-452. 1
section of the chapter provides a non-technical description of radar and of how that technology relates to the military capability selected as the focus for the thesis.
The fourth section of the chapter describes the broad metrics used in the thesis to frame a comparison of performance in exploiting a technology to provide novel solutions to requirements for military capability. The fifth section of the chapter describes the concept of ‘system’ used in the thesis to organise the enquiry on this basis. The sixth section describes the research method used in the thesis to identify the causal linkages within the system so identified. The seventh and final section of the chapter describes the structure of the thesis.
1.1 Thesis limits: Military innovation in small democracies The thesis focuses on democracies because the governance arrangements in democratic nation states require defence actors to account for their choices and associated resource allocations to the nation’s taxpayers. The need to make innovation choices legitimately constrains democratic defence actors in ways not encountered by their counterparts in non- democratic states. The thesis does not address military technological innovation in non- democratic states.
The thesis focuses on small democracies, defined as nation states with a population of less than 30 million and including, for example, Australia, Israel, Singapore, South Africa, Sweden and Switzerland. These and similar small democracies seek to preserve a prudent degree of military sovereignty but encounter resource constraints that require them to choose between indigenous ‘make’ and imported ‘buy’ solutions to their military capability requirements. Conversely, larger democracies (for example, the United States (US), Japan, Britain, France and Germany with populations ranging from 50-200 million) face less compelling resource constraints and enjoy greater latitude in choosing between ‘make’ and ‘buy’ solutions. This thesis does not address military technological innovation in larger democracies.
The research undertaken in this thesis in response to the above research question focuses on Sweden and Australia. In population terms these are two relatively small capitalist democracies with comparable gross domestic product( GDP)/capita. Both operate open, trading economies and each enjoys considerable discretion in choosing between domestically developed ‘make’ solutions and overseas sourced ‘buy’ solutions to military capability requirements. Finally, both Sweden and Australia publish, or permit the publication of, information about their respective military technological innovations in English.
2
1.2 Thesis focus: Military capability and military technological innovation This section analyses the reference in the above research question to solutions to requirements for military capability. In this thesis a ‘capability’ is a contingent capacity for humans to do something. What humans do is a function not only of the objective capacity of any artefact those humans employ, but also of what the humans involved understand about the artefact and what they think they can accomplish by using it. This cognitive dimension of capability becomes increasingly significant as the complexity of the artefact increases.
In this thesis, then, a military capability is a contingent capacity to achieve a military effect. ‘Military capability’ comprises force structure and preparedness. Force structure is the combination of artefacts, doctrine and people trained in the use of those artefacts in accordance with that doctrine. Preparedness is the capacity of a given force to generate a specified military effect by a stipulated time and to sustain that effect for a stipulated period of time. Preparedness is about the use of extant military capabilities. As such, preparedness generates little incentive for technological innovation but creates strong incentives for tactical and other innovation in the way in which, and the purposes for which, an extant capability is used. This thesis does not address preparedness-related innovation.
This clears the way for discussion of the reference in the research question to ‘novel solutions to requirements for military capability’. To provide such solutions, national defence actors must adjust the extant stock of national military capabilities through new investment. Such new investments may be prompted by external developments or by the need to replace obsolete artefacts so as to maintain extant military capability. A key impetus for such adjustments is the perceived need to adjust the force structure element of national military capability in order to address a national security problem. But a decision to adjust the stock of military capability is a necessary, not a sufficient, condition for military technological innovation.
Depending on the domestic resources available and on the extent to which they participate in the international division of labour, nations confront a spectrum of choices ranging from indigenous development and production (‘make’ solutions) to imports (‘buy’ solutions). Both ‘make’ and ‘buy’ solutions may entail substantial innovation. But to the extent that ‘make’ solutions entail the indigenous development of new technology or the novel application of existing technology, it is ‘make’ solutions that provide the prime impetus to military technological innovation. Hence in addressing the research question stated at the beginning of the chapter, this thesis focuses on ‘make’ solutions to capability requirements. The substantial organisational and doctrinal innovations often required to absorb an overseas-sourced ‘buy’ solution are outside the ambit of the thesis.
3
1.3 The thesis technology focus: Radar The outcome of the competition between antagonists for military advantage depends on, among many other factors, those antagonists’ respective situational awareness. The need to maintain such situational awareness generates a requirement for a broad area surveillance capability. Since World War Two, the small democracies with which this thesis is concerned have used various kinds of radar in providing solutions to their requirements for broad area surveillance capabilities.
Radar is based on electro-magnetic radiation, a phenomenon first hypothesised by Maxwell in the 1860s, demonstrated experimentally by Hertz in the 1880s and first exploited (for long-range communication purposes) by Marconi in the early 1900s. Hertz also showed that metal objects reflected electro-magnetic radiation. In 1904 Hulsmeyer drew on Hertz’s experiments in using electro-magnetic radiation to detect iron ships but his device failed commercially. In the late 1930s, however, Britain’s Air Ministry began developing a radar- based broad area surveillance capability as part of its air defence system. Britain succeeded in deploying this system in time to defeat the Luftwaffe’s daylight attacks in 1940. Thereafter, the exigencies of the Second World War caused radar technology to be developed extensively by, and to diffuse rapidly among, the countries involved.
The competition for military advantage causes military-related technology to diffuse from technology leaders to technology followers. The historical record suggests that efforts by technology leaders to obtain military advantage by keeping technological innovation secret or by otherwise denying adversaries access to technology serves to delay, but does not preclude, such diffusion. Radar is one such technology that has diffused widely, not only among nations but also between civil and military applications.
Both Sweden and Australia had similar requirements for a broad area surveillance capability. Both countries invested heavily in radar-based solutions to their respective requirements for a broad area surveillance capability during the latter half of the Cold War period. During this period both Sweden and Australia had comparable access to advanced US, British and other Western radar technology. Radar-based innovation is therefore an appropriate starting point for developing a comparison of innovation performance in Sweden and Australia. In order to establish a specific focus for this comparison, the thesis will compare and contrast Swedish development of the ERIEYE rapid reaction surveillance radar and the Australian development of the Jindalee Operational Radar Network (JORN).
1.4 Thesis metrics: Innovation performance This thesis compares Swedish and Australian innovation performance in terms of the efficiency and effectiveness with which each country adapted radar-based technology in meeting their respective requirements for a broad area surveillance capability. In order to gauge the efficiency with which each country adapted comparable radar technology to meet comparable requirements for a broad area surveillance capability, this thesis uses two 4
metrics. The first metric is the time each took to develop radar-based solutions to their respective requirements. The second metric is the cost they incurred in doing so.
In order to gauge the effectiveness with which each country adapted radar technology to meet their requirement for a broad area surveillance capability, this thesis uses one test and one metric. The test of effectiveness used in the thesis is whether or not the radar-based innovation was accepted into service by the Swedish or Australian user responsible for broad area surveillance. The metric of effectiveness used in the thesis is how, and the extent to which, that radar-based innovation diffused after it was accepted into service by the national user concerned.
1.5 Thesis framing: A system perspective The thesis uses a systems approach in framing its investigation of innovation performance in Sweden and Australia. According to Hall, a set of relationships can only comprise a ‘system’ if the relationships are somehow causally related to each other.2 Fleck extended Hall’s notion of causal connection in defining systems as “complexes of elements or components, which mutually condition and constrain one another, so that the whole complex works together, with some reasonably clearly defined overall function”.3 Fleck’s definition underpins the systems approach used in this thesis.
The research has also been informed by Hall’s distinction between mechanistic and evolutionary approaches to the explanation of change in economic systems. What Hall calls the mechanical approach to explaining change in economic systems assumes that the analyst can discover fundamental laws of nature which determine the state of the system at any time. The implication is that, if one knows the laws, one can predict what will happen next. By contrast, according to Hall, the evolutionists ‘look inside’ the system for the mechanisms generating change. Systemic change so identified is inherently path dependent. Evolutionary-based explanations of such change emphasise the historicity and contingent nature of such change. Hall echoed the earlier work of Nelson and Winter in arguing that the evolutionary perspective of innovation is inherently probabilistic, not deterministic.4 This thesis adopts Hall’s evolutionary perspective of system change. Hence the causal connections identified in the investigation of innovation performance in Sweden and Australia are probabilistic rather than deterministic.
This thesis has also been informed by Potts’s notion of a systems-based theory of evolutionary micro-economics. Particularly relevant for present purposes is Potts’s conception of economic systems in terms of elements (for example, economic agents, commodities, endowments) placed in some sort of relationship to each other. Such
2 Peter Hall, Innovation, Economics and Evolution, Harvester Wheatsheaf, London, 1994, p. 6. 3 James Fleck, Configurations: crystallizing contingency, International Journal of Human Factors in Manufacturing, Vol 3 (1), 1993, p. 17. 4 Hall, pp. 6-8. 5
connections are manifest in the structure of interdependence and interaction among economic agents, in the modalities of technology and in forms of organisation and competence. Also relevant for the purposes of this thesis is Potts’s emphasis on the bounded rationality of economic actors: Specifically, this thesis accepts Potts’s argument that, because economic actors are boundedly rational, economic systems do not constitute a single cohesive entity in which everything is connected to everything else. The thesis also accepts Potts’s argument that, however, once a system emerges as an entity, it can serve as a singular building block for a higher-level system.5 This notion of system-element duality becomes particularly important in the analysis of ERIEYE and JORN development later in the thesis.
1.6 Thesis research: Case study methodology In addressing the above research question this thesis uses a case study method. Yin defines a case study as an empirical enquiry that investigates a contemporary phenomenon in depth, within its real-life context and where the boundaries between that phenomenon and its wider context are not clearly evident.6 Because Yin’s case study method can be generalised to theoretical propositions, it is particularly well suited to the investigation undertaken in this thesis of causal relations embedded in military technological innovation.
Yin insisted, however, on theory development as part of the process of designing case study research.7 For the reasons discussed below, however, this thesis places emphasis on constructing a framework for a comparative analysis of cases studies rather than attempting to construct a theory to explain, ab initio, the causal relationships among the variables revealed by analysis of the case studies. The following paragraphs explain the approach to case study theory adopted in this thesis.
According to Eisenhardt and Graebner, “The theory is emergent in the sense that it is situated in and developed by recognising patterns of relationships among constructs within and across cases and their underlying logical arguments.”8 This notion of patterns of relationships underpins the research design adopted in this thesis. The research design also entails selecting case studies with a view to formulating analytical generalisations – what Eisenhardt called theoretical sampling.9 As Eisenhardt and Graebner emphasised, there is nothing random in the selection of cases for this purpose. Rather, a researcher will select those cases considered most likely to yield theoretical insight in terms of, for example, revelation of an unusual phenomenon, replication or contradiction of findings from other
5 Jason Potts, The New Evolutionary Microeconomics: Complexity, Competence and Adaptive Behaviour, Elgar, Cheltenham, UK, 2000, p. 2. 6 Robert K Yin, Case Study Research Design and Methods (Fourth Edition), Sage Publications, Thousand Oaks, California, 2009, p. 18. 7 ibid., pp. 35-40. 8 Kathleen Eisenhardt and M. Graebner, Theory building from cases: opportunities and challenges, Academy of Management Journal, Vol 50 (1),2007, p. 25. 9 Kathleen Eisenhardt, Building theories from case study research, Academy of Management Review, Vol 14 (4), 1989, p. 537. 6
cases, elimination of rival explanations and extension or elaboration of emergent theory.10 There is sufficient information on the public record about Swedish and Australian military technological innovation to sustain a theoretical sampling approach along the above lines.
Walsham placed more emphasis than Yin on preserving a considerable degree of openness to the field data and a willingness to modify initial assumptions and theories. Walsham emphasised the iterative process of data collection and analysis, with initial theories being modified or even abandoned in the light of empirical research.11 Walsham’s iterative approach accords closely with the constant comparison between emergent theory and data advocated by Eisenhardt with a view to iterating towards a theory which closely fits the data.12 This iterative approach to development of a theory closely related to case study data is well suited to addressing the research question stated at the beginning of this chapter. The iterative approach also informed the selection of Malerba’s sectoral systems of innovation model to frame the case studies – see Chapter 2.
According to Yin, robust case study research requires the case studies to exhibit construct validity, internal validity, external validity and reliability.13 In order to demonstrate construct validity this thesis uses multiple sources of evidence in assembling the case studies. The main sources of evidence are documents, including government reports, organisational reports and historical studies. In the case of the Swedish case studies, the task of establishing construct validity using documentation was made much easier by the practice of Swedish government agencies, companies and universities to publish such information in English. Secondary sources of evidence include interviews and information posted online.
In order to establish the internal validity of the case studies, the thesis developed a generic model of a defence innovation system. This generic model is based on Malerba’s sectoral system of innovation framework (see the literature review presented in Chapter 2 for more detail) but adapted for the specific requirements of analysing military technological innovation from a system perspective. A particular strength of the Malerba framework for present purposes is that it helps the researcher organise case study material but stops short of proposing a theory of cause and effect. This framework enables the case study research undertaken for this thesis to comply with the iterative approach advocated by Walsham, Eisenhardt and Graebner (see above) while simultaneously enabling the identification of matches and discrepancies advocated by Yin.
Yin’s reference to the external validity of case study research relates to the generalisability of the theoretical conclusions drawn from that research. To establish the generalisability of case study research results Yin advocates the use of replication logic in the same way that
10 Eisenhardt and Graebner, p. 27. 11 G. Walsham, Interpretative studies in IS research: nature and method, European Journal of Information Systems, Vol 4, 1995, p. 76. 12 Eisenhardt, p. 541. 13 Yin, pp. 40-45. 7
scientists repeat experiments under known conditions to verify the findings of a single experiment. In scientific practice, the more often an experiment conducted under the same conditions yields the same results, the more robust the conclusions that can be drawn about cause and effect. Similarly, the more an experiment yields divergent results for anticipatable reasons, the more robust the conclusions the analyst can draw about cause and effect.14
In order to establish a case study method that complies with Yin’s requirements for internal and external validity, the thesis uses a combination of what Yin calls holistic case studies and embedded case studies.15 The analysis of the structure and operation of the Swedish and Australian defence innovation systems constitute holistic case studies. In order to improve the internal validity of the Swedish and Australian case studies at system level, however, the thesis uses the development, production and procurement of technologically similar artefacts to establish a basis for comparing the performance of each system. These artefact- level studies constitute what Yin calls ‘embedded case studies’.
Specifically, in order to illustrate how the distinctive features of the Swedish innovation system affected Swedish defence innovation outcomes, the thesis analyses the time taken to develop the ERIEYE airborne microwave radar system, the cost incurred in doing so and the pattern of ERIEYE development/diffusion following its acceptance into Swedish Air Force (SwAF) Service. Similarly, in order to illustrate how the distinctive features of the Australian innovation system affected Australian defence innovation outcomes, the thesis analyses the time taken to develop the JORN ground-based high-frequency radar system, the cost incurred in doing so and the pattern of JORN development/diffusion following its acceptance into Australian Air Force Service. To summarise, the empirical research underpinning the thesis is based on a structured cross-case comparison of a combination of holistic system-level case studies and embedded artefact-level case studies. This comparison establishes a reasonably robust replicative logic that underpins a defensible claim as to the external validity and generalisability of the research results.
In order to establish the reliability of the above case study methodology the documents used are cited and the data they provide critically assessed. In addition, the records of the interviews of Swedish witnesses conducted for the Swedish case studies have been checked with those witnesses in order to verify the accuracy of the statements attributed to them. In the Australian case studies, the thesis has drawn heavily on the Hansard transcripts of a series of hearings conducted by a Parliamentary committee and on submissions made by witnesses to that committee. The Hansard transcripts and most of the submissions were released by the committee secretariat for the purposes of the present research and constitute a reliable record of the views of those participating in the hearings.
14 ibid., pp. 54-59. 15 ibid., pp. 59-60. 8
1.7 Thesis structure In order to address the research question posed at the beginning of this chapter, the thesis comprises framing, empirical and analytical components. The framing component establishes the notion of a defence sectoral innovation system. The empirical component presents the Swedish and Australian case studies. The analytical component compares and contrasts the performance of the Swedish and Australian defence sectoral innovation systems as demonstrated in the findings of the case studies.
In accordance with this structure, Chapter 1 of the thesis introduces the research question, explains the overall approach adopted in the thesis to address the research question and describes the research method used. Chapter 2 provides a critical overview of the defence innovation literature from the perspective of the research question and establishes the knowledge gap addressed in the thesis. Chapter 2 then provides a selective review of the wider innovation literature as it relates to the research question. Chapter 2 also explains the selection of Malerba’s sectoral systems of innovation framework as the starting point for developing a generic model of a defence innovation system that meets the need for internally valid case study research discussed above.
Chapter 3 is devoted to adapting Malerba’s generic sectoral systems of innovation framework for the purposes of investigating the Swedish and Australian case studies. Malerba structured his sectoral systems of innovation around a series of ‘building blocks’. These comprise knowledge, technology, demand, actors and networks and, finally, institutions. Chapter 3 re-orders, re-defines, amplifies and extends these generic building blocks in order to facilitate their use in analysing the conduct, structure and performance of defence sectoral innovation systems.
The empirical component of the thesis comprises six chapters, of which three are devoted to the Swedish case studies and three to the Australian case studies. Accordingly, Chapter 4 describes the structure of the Swedish defence sectoral innovation system. This description is structured around the generic framework based on the modified and re-ordered building blocks developed in Chapter 3. Chapter 5 describes the development, procurement and diffusion of the ERIEYE radar system to illustrate the operation of the Swedish defence sectoral innovation system. Chapter 6 analyses how the distinctive features of the Swedish defence sectoral innovation system affected the time taken to develop ERIEYE, the cost incurred in doing so and the pattern of ERIEYE development/diffusion following its acceptance into service by the SwAF.
The Australian case studies are presented in the same way. Accordingly, Chapter 7 describes the structure and operation of the Australian defence sectoral innovation system. Chapter 8 describes the development, procurement and diffusion of JORN. Chapter 9 explains how the distinctive features of the Australian defence innovation system affected the time taken to
9
develop JORN, the cost incurred in doing so and the pattern of JORN development/diffusion following its acceptance into service by the Australian Air Force.
Chapter 10 concludes the thesis and, to this end, is divided into two sections. The first section is devoted to comparing and contrasting the performance of the Swedish and Australian defence innovation systems in terms of the time taken to develop the ERIEYE and JORN radar systems, the cost incurred in doing so and the pattern of development/diffusion following their respective acceptance into service. The second section summarises how a modified version of Malerba’s sectoral system of innovation model was used to fill an identified gap in the literature on military technological innovation. The second section also draws attention to insights gained in the process of adapting the sectoral system of innovation model for use in analysing military technological innovation that suggest directions for future refinement of that model. The second section concludes with observations about what the model of defence sectoral innovation system developed in this thesis suggests for the development and application of future defence innovation policy in small democracies.
10
Chapter 2: Literature Review
This chapter reviews selected literature relating to the question: why do similar countries addressing comparable requirements for military capability differ in terms of the time they take to meet those requirements, the cost they incur in doing so and the post-acceptance pattern of diffusion/development displayed by the solutions to those requirements? The review begins with a brief and very selective discussion of that literature dealing with the definition of innovation in general and of military technological innovation in particular. The chapter then reviews the literature on military technological innovation in terms of what light it sheds on the research question addressed in this thesis. The chapter then discusses the literature on national systems of innovation, technological systems of innovation, geographical systems of innovation and sectoral systems of innovation.
2.1 Defining innovation and military technological innovation In his overview of the innovation literature, Fagerberg distinguished between invention (defined as the first occurrence of an idea for a new product or process) and innovation defined as the first attempt – which often lags invention by a considerable period of time – to carry out an idea in practice.16 Fagerberg’s definition of invention is appropriate for present purposes. As the reference in Chapter 1 to Hulsmeyer’s attempt to use electro- magnetic radiation to detect ships suggests, however, Fagerberg’s definition of innovation needs refinement in the present context. Specifically, Fagerberg understates the technological preconditions, effort and ingenuity generally required to realise an idea in practice.
Accordingly, in this thesis the term innovation is used to denote the successful exploitation of an idea to perform a new function or to perform an existing function better. Here success is simply the commercialisation of an innovation or, in the present context, its acceptance into military service. This notion of successful exploitation of an idea was developed further by Tornatzky and Fleischer in their discussion of technological innovation.17 Zarzecki extended the latter concept to military technological innovation which he defined in terms of “any situationally new development and deployment of any knowledge-derived tool, artefact, or device used by the military forces of a state in the conduct of, or preparation for, armed conflict”.18
16 J. Fagerberg, Innovation – a guide to the literature in J. Fagerberg, D. Mowery and R. Wilson (eds): The Oxford Handbook of Innovation, Oxford University Press, Oxford, 2005, pp. 4-5. 17 Louis G. Tornatzky and Mitchell Fleischer, The Process of Technological Innovation, Lexington Books, Lexington, 1990, p. 11. 18 Thomas Zarzecki: Arms Diffusion: The Spread of Military Innovation in the International System, Routledge, New York, 2002, p. 74. 11
Zarzecki’s reference to ‘situationally new’ tools and technology highlights an important aspect of the present research. In this thesis the term ‘situationally new’ describes a situation in which an artefact is new to the entity contemplating its adoption. As Zarzecki put it: “a state is said to ‘innovate’ when it accepts for use a piece of military technology which it itself has never before adopted. This is true regardless of if or when the innovation was adopted by other states.”19
2.2 The defence innovation literature According to Zarzecki , any scheme for making sense of military technological innovation should be, firstly, inclusive in the sense that the scheme should allow for the inclusion of the broad spectrum of military technologies, ranging from the simplest to the most complex. Secondly, such a scheme should be complex in the sense that it can meaningfully reflect different levels of change in military technology over time. Finally, such a scheme should be relevant in the sense that it can create categories that accommodate the various political, economic and technological forces that can reasonably be expected to shape military technological innovation.20
In the following paragraphs, Zarzecki’s schemata is used to frame a review of several categories of military technological innovation literature. The first is histories of military technological innovation. The second category is the defence capability development literature, and particularly that sub-category dealing with the management of defence research and development. The defence economics literature constitutes a third category.
A major category of the military technological innovation literature comprises histories of specific military technological innovations. The starting point for this section is Hacker’s encyclopaedic review of the literature on the history of military technology.21 In his review, Hacker traversed what he calls the traditional focus on military hardware and then drew attention to an emerging theme in the literature dealing with military technology and society – the so-called new history of military technology. In canvassing the main themes of this new history Hacker mentioned studies of the interactions between military and economic institutions mediated through technology under varying political regimes.22 McNeill wrote what is arguably the exemplar of these socio-economic histories of military technology.23 Such histories, however, tend to focus on the consequences of the diffusion of various military technologies and place less emphasis on how they originated in the first place. In addition, they are too general to provide a basis for a study of innovation performance, the focus of this thesis.
19 ibid., pp. 74-75. 20 ibid., p. 78. 21 Barton C.Hacker, Military institutions, weapons, and social change: towards a new history of military technology, Technology and Culture, Vol 35 (4), October 1994, pp. 768-834. 22 ibid., pp. 823-824. 23 William McNeill,:The Pursuit of Power – Technology, Armed Force, and Society since AD 1000, The University of Chicago Press, Chicago, 1982. 12
As indicated in Chapter 1, this thesis focuses on case studies of radar-related innovation in Sweden and Australia. Hence histories of radar technology might be seen as a logical starting point for a review of the literature pertinent to this thesis. Representative examples of such technological histories include Zimmerman’s history of Britain’s development of radar in the lead up to World War Two24 and Allison’s history of the US Navy’s development of radar in the same period.25 Brown complements these in-depth studies with his historical survey of essentially concurrent radar development undertaken in the US, Britain, Germany, the Soviet Union, Japan and France in the lead up to and during World War Two.26
These histories are an invaluable source of the empirical material used to illustrate conceptual points made in the thesis. By definition, however, such histories are idiosyncratic. As such, they fail Zarzecki’s inclusiveness test. In addition, while such histories corroborate the proposition that countries at comparable stages of economic development faced with comparable problems may embark on comparable development programs, the histories shed little light on why their performance varies so widely.
A similar point applies to histories of the diffusion of military technology. For example, Zimmerman’s study of the exchange of radar technology between Britain and the US during World War Two provides a compelling account of how, given the appropriate policy environment, military technological innovation diffuses from technology leader to technology follower, is absorbed by the latter who then modifies it, after which it diffuses back to the original technology leader in enhanced form.27 Similarly, Mellor described how British radar technology diffused to Australia in the very early stages of World War Two and how Australians adapted that technology to jungle warfare in the Pacific theatre.28 A more contemporary example is Zarzecki’s analysis of how artefacts incorporating military technological innovation diffuse through the international system.29 Similarly, the information technology based Revolution in Military Affairs (RMA) prompted Goldman and Eliason to edit a book exploring the modalities by which ideas like the RMA and the artefacts and doctrines incorporating those ideas diffuse internationally.30 The general point is that, while such histories of technological diffusion demonstrate convincingly that different countries absorb and adapt comparable technologies at different rates and to varying extents, they shed little light on why those differences occur.
24 David Zimmerman, Britain’s Shield: Radar and the Defeat of the Luftwaffe, Sutton Publishing, Stroud, UK, 2001. 25 David K. Allison, New Eye for the Navy: The Origin of Radar at the Naval Research Laboratory (NRL Report 8466), Naval Research Laboratory, Washington D.C., 29 September 1981. 26 Louis Brown, A Radar History of World War II, Institute of Physics Publishing, Bristol, 1999. 27 David Zimmerman, Top Secret Exchange: The Tizard Mission and the Scientific War, McGill-Queen’s University Press, Montreal & Kingston, 1996. 28 D.P. Mellor, The Role of Science and Industry, Australian War Memorial, Canberra, 1958, pp. 423-452. 29 Zarzecki. 30 Emily Goldman and Leslie Eliason (eds): The Diffusion of Military Technology and Ideas, Stanford University Press, Stanford, California, 2003. 13
Similar observations can be made about histories of radar’s role as a disruptive technology. For example, Dobinson described how, in the late 1930s, radar displaced sonar-based technologies as a means of providing early detection of incoming aircraft in Britain.31 This kind of history is relevant in the sense that it describes in rich detail the political, technological and organisational developments that characterised the displacement. But it does not provide a framework that can be generalised to other jurisdictions and other technologies.
The documentation of military technological developments sponsored by various elements of the US defence community constitutes a separate literature relevant to this thesis. These studies, often undertaken by the US National Defense University, are typically more normative than the histories discussed above. Their primary relevance to the present research is what they indicate about efforts to use historical study to inform contemporary policy activity.
One landmark example of this sub-set of the military innovation literature is Project Hindsight, a report commissioned in the late 1960s by the US Defense Department’s Director of Defense Research and Engineering (USDDR&E) in response to concerns about the cost-effectiveness of US Defense research and development spending. Project Hindsight involved a retrospective examination of some 20 US-developed weapon systems (including missiles, torpedoes, radars, night vision equipment, transport aircraft and nuclear warheads) with a view to, firstly, gauging the payoff to the US Defense Department of its own investment in science and technology and, secondly, identifying patterns of management that might improve US defence innovation outcomes.32 In 2005 a team of US National Defense University (NDU) scholars used the 1967 Project Hindsight report as a point of departure for a study of US Army weapon system developments including the Abrams tank, Apache helicopter gunship, Stinger and Javelin missile systems.33
Taken together, the 1967 Project Hindsight report and its 2005 sequel constitute a useful longitudinal study of military technological innovation. Particularly relevant for present purposes are the continuities and discontinuities in the development and fielding of military technology highlighted in the studies. For example, the two studies highlighted the enduring importance of basic scientific and technological research into observed phenomena as foundation for development work for the systems of interest. This continuity informed the case studies developed in the thesis. The two studies also drew attention to the recent use
31 C. Dobinson, Building Radar, Methuen, London, UK, 2010. 32 Chalmers Sherwin and Raymond Isenson, Project Hindsight: a Defense Department study of the utility of research, Science, Vol 156, June 1967, pp. 1571-1577, available at http://www.sciencemag.org/content, accessed 3 May 2011. 33 John Lyons, Richard Chait and Duncan Long, Critical Technology Events in the Development of Selected Army Weapons Systems: A Summary of Project Hindsight Revisited, Centre for Technology and National Security Policy, National; Defense University, Washington D.C., September 2006, available at http://www.ndu.edu/ctnsp/publications.html, accessed 3 June 2011. 14
of modelling and simulation techniques in the development and fielding of military technology. This discontinuity also informed the case studies developed in the thesis.
Equally relevant for present purposes, however, are the methodological deficiencies that characterised these studies. These deficiencies hampered the generalisation of the studies’ conclusions. In order to avoid a similar methodological weakness, the present research pays much more attention to selecting and articulating a robust case study methodology (see Chapter 3).
Preparation of this thesis was also informed by another, broader NDU study, also published in 2005 but undertaken with a view to ascertaining how science and technological innovations occur.34 This study reviewed the development of early radar and then other innovations involving several scientific disciplines, including solid state digital electronics; cellular telephone systems; the global positioning system; DNA fingerprinting; compact disk and digital versatile disk technology; and giant magneto-resistance computer memory read heads.
This NDU study concluded that science-based technological innovation typically begins with an early searching phase (the ‘prospecting phase’) characterised by a few discrete but high- impact events typically associated with the work of individuals. This is followed by a later, more predictable phase (the ‘mining phase’) dominated by continuous improvement in functional capability of the innovation through group-based research.35
For present purposes, the most immediately relevant aspect of the NDU study was its distinction between the prospecting and mining phases of military technological innovation, its corroboration of the need to take into account the role of the individual actor as well as its convincing demonstration of the importance of like-minded groups in progressing such innovation. Methodologically, the NDU study focused on a broad spectrum of technologies undertaken within the US national innovation system. This leaves room for a complementary approach (adopted in this thesis) which focuses on deep, rich case studies of a single technology (radar in the present case) but undertaken in diverse jurisdictions (Sweden and Australia).
A related but divergent body of literature on the history of military technological development concentrates less on specific technologies and more on organisational and institutional arrangements by which technologies were developed. For example, Lassman’s investigation of the sources of US weapon systems innovation concentrates on the US Defense Department’s internal research and development activities, with particular reference to the patterns of organisational change that guided the development of major
34 Timothy Coffey, J. Dahlberg and Eli Zimet, The S&T Conundrum, August 2005, National Defense University, available at http://www.ndu.edu/CTNSP, accessed 14 June 2011. 35 ibid., p. 15. 15
weapon systems.36 The emphasis on organisational arrangements and institutional processes in this category of US historical studies is of general interest to the present research. However, such US-centric explanations of the various political, economic and technological forces that shape military technological innovation need considerable qualification before they can be applied to analysis of military technological innovation of the small democracies with which the thesis is concerned.
The need for a less US-centric approach was at least partly met by the scholarly response to the United Kingdom (UK) government’s privatisation of the British defence research laboratories in the 1990s and efforts by members of the European Union (EU) to establish workable intra-European arrangements for collaborative military research and development. These initiatives prompted a considerable amount of case study based research on the relationship between organisational and institutional arrangements and military technological innovation. Particularly relevant for this thesis is the work by, for example, James37 and Molas-Gallart38 who applied a systems perspective to their analysis. This thesis seeks to complement that work by undertaking an in-depth study of two national defence innovation systems focused on a single broad technology.
Alic’s study of US military technological innovation is particularly instructive in this regard. The approach adopted in this thesis was informed by Alic’s nuanced treatment of the disparate activities of technological innovation, warfighting practices and acquisitions politics and policy that characterise US military technological innovation.39 Despite its US focus, Alic’s analysis of, for example, innovation and learning has much to offer to studies of military technological innovation in small democracies. However, Alic’s analysis ranges broadly over numerous technologies and disparate timeframes and, as such, complements the narrower but more detailed analysis attempted in this thesis.
What might be termed the defence programming literature constitutes a second, less historically oriented category of literature that addresses defence technological innovation in the context of military capability development. The classic example of this genre is Hitch and McLean’s book on defence planning and budgeting in the US.40 Particularly relevant for present purposes is the way this and similar writing conceives strategy, technology and
36 Thomas C. Lassman, Sources of Weapon Systems Innovation in the Department of Defense: The Role of In- house Research and Development, 1945-2000, Centre of Military History, United States Army, Washington D.C. 2008. 37 See, for example, Andrew D. James, Organisational change and innovation system dynamics: the reform of the UK government defence research establishments, Journal of Technology Transfer, Vol 34, 2009, pp. 505- 523. 38 Notably Jordi Molas-Gallart, Government defence research establishments: the uncertain outcome of institutional change, Defence and Peace Economics, Vol 12 (5), 2001, pp. 417-437. 39 John A. Alic, Trillions for Military Technology – How the Pentagon Innovates and Why It Costs So Much, Palgrave, New York, 2007. 40 Charles J. Hitch and Roland N. McLean, The Economics of Defense in the Nuclear Age, Harvard University Press, Cambridge, 1967. 16
economy as interdependent elements of the same problem rather than as three independent considerations to be assigned appropriate weights. In this sense strategies are ways of using budgets or resources to achieve military objectives and technology defines possible strategies. The need for defence research and development derives from the interaction between changing military objectives and technological limits. The resulting investment in research and development is planned as part of an economic calculus.
The economic perspective of defence research and development propounded by Hitch and McLean is implicitly neo-classical. This orientation understates the importance to defence innovation outcomes of the bounded rationality of the defence actors involved. This deficiency undermines the analytical utility of the Hitch and McLean model in achieving a major objective of this thesis – that is, helping both scholars and practitioners understand how and why military technology changes over time.
The defence economics literature constitutes a third major category of literature in which defence technological innovation is addressed. A prominent example is Hartley and Sandler’s foundational text of defence economics literature.41 This is complemented by numerous articles published in the journal Defence and Peace Economics. This literature tends to focus on such themes as the impact of military expenditure on economic growth, arms races and arms proliferation, the defence industrial base and the economics of defence research and development (R&D). But this literature sheds surprisingly little light on the research question that this thesis addresses. This is primarily because it tends to focus on the various political, economic and technological forces that account for military capability development at a particular point in time at the expense of identifying and explaining those forces that shape the outcomes of military technological innovation over time.
A fourth category of defence innovation literature focuses on the development of advanced military capabilities but instead of focusing on the political economy of Pentagon research and development it focuses on the interaction between national resources and military capability. Smith’s study of military economics is a good example of this category.42 This voluminous category tends to be characterised by a strong normative thrust and to have a more or less clear policy edge and motivation. This kind of literature can inform policymaking by, for example, illustrating various dimensions of the capability development process in considerable detail. But the military capability development literature tends to be less complex in the sense that it does not probe the subtlety of the military innovation process to the extent required for the purposes of this thesis. The military capability development literature also tends to be less relevant to the purposes of this thesis in that the literature tends to frame consideration of political, economic and technological
41 Keith Hartley and Todd Sandler: Handbook of Defense Economics, Elsevier, 1995. 42 Ron Smith, Military Economics – The Interaction of Power and Money, Palgrave, New York, 2009. 17
influences on military technological innovation in terms of a single nation’s institutional arrangements (often those of the US or other larger nations) and of a single author’s policy predilections.
In order to rectify the deficiencies of the defence innovation literature, the next section considers selected aspects of the broader innovation literature.
2.3 The wider innovation literature The primary focus of this section is a review of the literature based on a systems perspective of the innovation process. To prepare the way, the following paragraphs survey selected writing about, firstly, the role of market demand and technological opportunity in the innovation process; secondly, the role of firms and other actors in the innovation process; and, thirdly, the diffusion and absorption of technology.
A distinctive feature of defence business in nation states is the role played by the national defence organisation as a monopsonistic buyer of bona fide defence goods and services, including military technological innovation. At issue, however, is whether such demand is a sufficient condition to explain observed outcomes of the military technological process. This thesis takes into account the analysis by Mowery and Rosenberg of the relative influence of market demand and of technological opportunity on innovation outcomes.43 The present research has been informed by their conclusion that, “Rather than viewing either the existence of a market demand or the existence of a technological opportunity as each representing a sufficient condition for innovation to occur, one should consider them each as necessary, but not sufficient, for innovation to result; both must exist simultaneously.”44
Metcalfe places more emphasis on the role of purposeful actors in bringing together both market demand and technological opportunity in the innovation process:
Technologies are articulated by purposeful organisations capable of search activity and capable of reacting, although often erroneously and within limits, to unanticipated events. There are plausible arguments for claiming that the nature and timing of inventions are random events but, equally, there are powerful inducement mechanisms at work in shaping the role and direction of inventive activity. Certainly the transition from invention to innovation is guided by selective forces. In terms of evolutionary theory, there is a clear Lamarckian element to be incorporated here. Not only do innovations arise in response to perceived needs and opportunities, they are carried
43 D. Mowery and N. Rosenberg: The influence of market demand upon innovation; a critical review of some recent empirical studies, in N. Rosenberg (ed.): Inside the Black Box: Technology and Economics, Cambridge University Press, Cambridge, 1982, pp. 193-241. 44 ibid., p. 231. 18
forward through time in the memory of firms and other institutions in such a way that the experience of the past shapes what they can achieve in the future.45
The bounded rationality of defence innovation actors is a prominent theme in this thesis. But Metcalfe’s reference to the Lamarckian search by actors that while boundedly rational are nevertheless purposeful, highlights the importance of search activity to innovation outcomes generally, and to defence innovation outcomes particularly. This leads to discussion of the locus of innovation activity.
The generic innovation literature also places firms at the centre of the innovation process. Nelson and Winter, for example, placed firms, and firms’ ‘routines’ (that is, regular and predictable behaviour patterns and strategic heuristics) at the centre of their study of innovation and long-run economic and social change, noting that “it is … an institutional fact of life that in Western market economies … much technical advance results from profit- oriented investment on the part of business firms.”46 Metcalfe took this logic further: “The fact that firms learn, have memory and possess mechanisms for maintaining memory over time in the face of changes in personnel is the source of the chief elements of irreversibility in the pattern of economic progress.”47 This thesis accepts that firms play a central role in military technological innovation in small capitalist democracies. However, the systems perspective discussed below suggests strongly that firms are only one actor in the process of military technological innovation.
Von Hippel’s investigation of the variations in the functional sources of innovation in a range of industries sheds a different light on the role of firms in the innovation process. In some cases, like scientific instruments, he observed that major product innovations were generally initiated by product users; in other cases, like earth-moving equipment, the manufacturer was the main source of innovation; and in still other cases – industrial gases and thermoplastics, for example – component suppliers were major sources of innovation.48 The complexity of military artefacts and the diversity of components, sub-systems and systems comprising them suggests that military technological innovation is likely to show a similar variation in functional sources.
Military technology diffuses among nation states. Rogers proposed a socio-centric model to describe the diffusion of technology from leaders to followers.49 The usefulness of Rogers’ model to this thesis is weakened by his abstraction from the specific characteristics of both the technology and the actors involved. In addition, the centre-periphery model of
45 J. S. Metcalfe, Evolution and economic change, in Ulrich Witt (ed.), Evolutionary Economics , Edward Elgar, Aldershot, UK, 1993, p. 370. 46 R. Nelson and S. Winter, An Evolutionary Theory of Economic Change, Harvard University Press, Cambridge, 1982, p. 28. 47 Metcalfe, p. 370. 48 Eric Von Hippel, The Sources of Innovation, Oxford University Press, Oxford, 1988, pp. 3-5. 49 Everett M. Rogers, Diffusion of Innovation (Fourth Edition), The Free Press, NY, 1995, p. 11. 19
communication that underpins his model is useful in explaining the deployment of discrete, unequivocal, easily valued and clearly meritorious technologies. But Rogers’ model is less appropriate where, as is the case in the military context, the technology is advanced and where technological innovation is characterised by developer and user iterating and shaping the idea of the tool and its possible uses.
The present research has also been influenced by the wider innovation literature’s treatment of a recipient’s capacity to absorb technology diffused along the above lines. Specifically, Westney argued against analysing cross-societal emulation on the basis of distinctions between copying and inventing and between imitation and innovation. All organisations must draw on the surrounding environment for resources and must respond to the external demand for their goods and services. Since the environment in which the organisational model was anchored is in its original setting will inevitably differ from the one to which it was transplanted, even the most assiduous emulation will result in alterations of the original pattern in order to adjust them to their new context. The present investigation of the diffusion and absorption of military technology has been informed by Westney’s observation that, “Some of these changes are deliberate; some are unintended, and virtually all will have unforeseen consequences. For some organisations, the original model will continue to provide a blueprint for development; for others, the original model will quickly lose ground to more powerful influences in the immediate environment.”50
The above highly selective and necessarily cursory survey of the wider innovation literature highlights the diversity of generic innovation concepts that need to be taken into account in answering the research question underpinning this thesis. This conceptual diversity highlights the need for an analytical framework that helps bring these and other issues into a coherent focus on the military technological phenomenon. As Fagerberg observed:
A central finding in the innovation literature is that a firm does not innovate in isolation, but depends on extensive interaction with its environment. Various concepts have been introduced to enhance our understanding of this phenomenon, most of them including the terms ‘system’ or (somewhat less ambitious) ‘network’. Some of these have become popular among policy makers, who have been constrained in their ability to act by lack of a sufficiently developed framework for the design and evaluation of policy. Still, it is a long way from pointing to the systemic character of innovation processes (at different levels of analysis), to having an approach that is sufficiently developed to allow for systematic analysis and assessment of policy issues. Arguably, to be really helpful in that regard, these system approaches are in need of substantial elaboration and refinement …51
50 Eleanor D. Westney, Imitation and Innovation: The Transfer of Western Organisational Patterns to Meiji Japan, Havard University Press, Cambridge, 1987, pp. 6-7. 51 Fagerberg, p. 20. 20
2.4 The innovation systems literature As Fagerberg indicated, the systems of innovation literature is a relatively recent phenomenon. That said, however, the systems of innovation literature has grown rapidly. This thesis draws heavily on certain elements of the systems of innovation literature. These elements are summarised in the following paragraphs.
Edquist traced the emergence and development of the systems of innovation approach. In doing so he conceived innovation not only in terms of the more familiar product innovation (new or better goods and services), but also as encompassing process innovations (new ways of producing goods and services). This inclusive perspective led him to envisage systems of innovation as encompassing all economic, social, political, organisational, institutional and other factors that significantly influence the development, diffusion and use of innovations.52
On this basis Edquist identified several variants of the systems of innovation approach which are discussed in the following paragraphs. The following overview begins with the technological systems approach which focuses mainly on networks of agents interacting in the generation, diffusion and use of technologies. This is followed by an overview of the national system of innovation approach which highlights the importance of national boundaries, non-firm actors (including government) and institutions specific to a given nation state in the development and diffusion of new technologies. An overview of the sub- national or regional system of innovation approach, focusing on geographic clusters of industries follows next. The overview concludes with an analysis of the sectoral system of innovation approach which complements rather than displaces the above variants of the systems of innovation approach and focuses on the firms engaged in developing and producing a sector’s goods and services and in generating and exploiting a sector’s technologies.53
Carlsson defined technological systems as “knowledge and competence networks supporting the development, diffusion and utilisation of technology in established or emerging fields of economic activity.”54 Carlsson’s notion of technological systems focuses on generic technologies (rather than industries) and comprises networks of firms, research and development infrastructures, educational institutions and policymaking bodies. One country can host several technological systems which, however, may transcend national borders.
52 Charles Edquist, Systems of innovation: perspectives and challenges in Fagerberg, Mowery and Wilson (eds), The Oxford Handbook of Innovation, Oxford University Press, Oxford, 2006, p. 182. 53 Charles Edquist: Systems of Innovation Approaches – Their Emergence and Characteristics in Charles Edquist (ed.), Systems of Innovation: Technologies, Institutions, and Organisations, Pinter, London, 1997, especially pp. 11-29. 54 Bo Carlsson, Technological Systems and Industrial Dynamics, Kluwer Academic Publishers, London, 1997, p. 2. 21
Military artefacts – particularly the sophisticated military artefacts with which this thesis is concerned – will represent a synthesis of inputs from several such technological systems. A modern combat aircraft, for example, will involve a host of aircraft platform technologies (including airframe and propulsion technologies), electronically based sensor technologies, electronically based communication technologies and electronically based data processing technologies, to name a few. Hence the technological systems approach facilitates deep analysis of certain elements of the military technological innovation phenomena but is poorly suited to providing the kind of integrated perspective required to inform the investigation of innovation performance intended in this thesis.
Within these limits, Granberg’s analysis of Sweden’s powder technology as a technological system illustrates how the technological system approach can inform analysis of military technological innovation. Granberg conceptualised technology in cognitive terms, that is, as bodies of knowledge and knowledge-generating processes rather than as bodies of artefacts. On this basis he defines powder technology as a field of technical knowledge and competence underlying the development and production of powders and powder-based materials for engineering applications requiring advanced structural materials properties. Granberg then focused on problem areas and competencies. This focus includes determining the means (tools, devices or technical systems) by which particular functions can be performed effectively and efficiently. It also involves determining where technology and technical knowledge grow by the identification, specification and solving of technical problems.
Granberg then identified major categories of actors, including policy and supporting organisations, academic departments, research institutes, suppliers of powder, suppliers of process equipment and chemicals, suppliers of materials and components, and users/system suppliers. The final step is to identify the basic features and general patterns of the systemic links – the network – linking first echelon users, the materials/parts suppliers and the infrastructure actors.
Such networks comprise overlapping and interrelated elements including an economic network (commercial actors connected by buyer-supplier relations). A key network involves technical problem solving (the focus of knowledge and competency flows). This is complemented by an extensive, informal national community network based on a common interest in development of the field, the nodes of which are formed by the leading practitioners.55 The technological systems approach focuses on technological problem solving as a prime engine of technological system activity. This approach is particularly instructive for the purposes of this thesis. Bergek and her colleagues extended this logic in their analysis of the functional dynamics of technological innovation systems. They defined
55 Anders Granberg, Mapping the cognitive and institutional structures of an evolving advanced materials field: the case of powder technology, in Carlsson, Technological Systems, pp. 169-200. 22
the latter in terms of a socio-technical system focused on the development, diffusion and use of a particular technology (in terms of knowledge, product or both).56 That said, however, the technological innovation system model is not sufficiently inclusive to provide a satisfactory framework for analysis of military technological innovation.
This leads to consideration of the national systems of innovation approach pursued by, among others, Freeman,57 Lundval58 and Nelson.59 For example, Lundval argued that national systems reflect basic differences in historical experience, language and culture which are manifest in the internal organisation of firms, in inter-firm relationships and in the role played by the public sector. They also argued that these differences are also reflected in the institutional set-up of each nation’s financial sector as well as in the intensity and organisation of R&D.60
Metcalfe’s analysis of national systems of innovation pulled these various themes together in a way that is particularly useful for the purposes of this thesis. Metcalfe defined a national system of innovation as: that set of distinct institutions which jointly and individually contribute to the development and diffusion of new technologies and which provide the framework within which governments form and implement policies to influence the innovative process.61
Metcalfe emphasised the role of interconnected institutions in national systems of innovation in the creation, storage and transfer of knowledge, skills and artefacts. According to Metcalfe, the element of nationality follows not only from the focus on technology policy (see below), but also from “elements of shared language and culture which bind the system together and form the national focus of other policies, laws and regulations which condition the innovative environment.”62 This thesis draws heavily on the national systems of innovation approach because, as a core public good, national defence is a central policy preoccupation for nation states in general and particularly for the small democracies on which this research focuses. A central premise underpinning this research is that defence policymaking, and within that framework, defence technological choices are products of national systems but draw on technological systems that transcend national boundaries.
56 A. Bergek, S. Jacobsson, B. Carlsson, S. Lindmark and A. Rickne: Analyzing rhe functional dynamics of technological innovation systems: a scheme of analysis, Research Policy, No 37, 2008, pp. 407-429. 57 C. Freeman, Technology Policy and Economic Performance – Lessons from Japan, Pinter, London 1987. 58 Bengt-Ake Lundval (ed.), National Systems of Innovation: Towards a Theory of Innovation and Interactive Learning, Pinter, London, 1992. 59 R. Nelson, National Systems of Innovation: A Comparative Study, Oxford University Press, Oxford, 1993. 60 Lundval, p. 13. 61 J.S. Metcalfe, The economic foundations of technology policy: equilibrium and evolutionary perspectives in P. Stoneman (ed.) Handbook of the Economics of Innovation and Technological Change, Blackwell Publishers, Oxford, 1995, p. 462. 62 ibid., p. 463. 23
There is an extensive literature focused on sub-national or regional systems of innovation or industrial development. A prominent example is Saxenian’s analysis of industry clusters in the Silicon Valley area of California and along Route 128 in Massachusetts.63 Saxenian’s characterisation of the difference between these two regions is applicable more generally. She noted that Silicon Valley has a regional network-based industrial system that promotes collective learning and flexible adjustment among specialised producers of a complex of related technologies. The region has dense social networks and open labour markets which, in her view, encourage experimentation and entrepreneurship. Her description of company behaviour in Silicon Valley is particularly instructive for this thesis. According to Saxenian, companies compete intensely while at the same time learning from each other about changing markets and technologies through informal communication and collaborative practices. Saxenian drew attention to loosely linked team structures which encourage horizontal communication among firm divisions and with outside suppliers and customers. The functional boundaries within firms are porous in a network system, as are the boundaries between firms themselves and between firms and local institutions, such as trade associations and universities.
Also instructive is Saxenian’s characterisation of the Route 128 region as. being dominated by a small number of independent, relatively integrated corporations that internalise a wide range of productive activities. Practices of secrecy and corporate loyalty govern relations between firms and their customers, suppliers and competitors, reinforcing a regional culture that encourages stability and self-reliance. Corporate hierarchies ensure that authority remains centralised and information tends to flow vertically. The boundaries within and between firms and between firms and local institutions thus remain far more distinct in this independent firm-based system.64
Overall, however, the geographical systems perspective of innovation is not sufficiently inclusive to provide a satisfactory framework for analysis of military technological innovation. However, Saxenian’s depiction of two distinctive innovation cultures is particularly instructive for present purposes. Saxenian also corroborated the observation by Breschi and Malerba to the effect that new knowledge, including that required for innovation, occurs more efficiently among closely located actors. A related theme, according to Breschi and Malerba, is that learning through networking and by interacting encourages firms to form clusters.65 In investigating military technological innovation, however, this thesis will distinguish between geographic clusters and functional clusters. In the latter case the research will investigate how innovation outcomes are influenced by the immersion of local firms in a very dense network of knowledge sharing which is supported by close social
63 A. Saxenian, Regional Advantage; Culture and Competition in Silicon Valley and Route 128, Harvard University Press, Cambridge, 1994. 64 ibid., pp. 2-3. 65 Stefan Breschi and Franco Malerba: Clusters, Networks and Innovation, Oxford University Press, Oxford, 2005, pp. 1-4. 24
interactions and by institutions that build trust and encourage informal relations among actors.
The sectoral systems of innovation approach is the fourth variant of the innovation systems approach identified by Edquist. Malerba, the main proponent of this approach, defined a sector as “a set of activities which are unified by some related product groups for a given or emerging demand and characterised by a common knowledge base”.66 Malerba extended this notion of a sector to a sectoral system of innovation and production which he defined as “a set of new and established products for specific uses and a set of agents carrying out activities and market and non-market interactions for the creation, production and sale of their products.”67
As Edquist observed68 – and as Malerba has acknowledged69 – the sectoral systems of innovation approach is not a theory but does provide a conceptual framework which facilitates descriptive analysis of sectors. The framework comprises the following building blocks:
• Knowledge
• Technologies
• Demand
• Actors and networks
• Institutions.70
As Malerba stressed, the sectoral system of innovation approach complements, rather than displaces, the variants of the systems of innovation approach discussed above. A particularly attractive feature of Malerba’s sectoral innovation systems approach for present purposes is its ability to accommodate those aspects of the other variants of the systems of innovation approach that inform present investigation of military technological innovation phenomena. That said, the sectoral systems of innovation approach needs amplification and adaptation if it is to provide a satisfactory approach to developing answers to the research question underpinning this thesis.
McElvey indicated a prospective direction for such amplification and adaptation. She suggested amplifying the approach by including a discussion of the mechanisms for the
66 Franco Malerba: Sectoral Systems of Innovation: Concepts, Issues and Analyses of Six Major Sectors in Europe; Cambridge University Press, Cambridge, 2004, p9. 67 ibid., pp. 15-17. 68 Charles Edquist, Systems of Innovation: Technologies, Institutions, and Organisations, Pinter, London, 1997, p. 28. 69 Malerba, p. 35. 70 ibid., pp. 17-29. 25
retention and transmission of information. She suggested extending the approach to include an explanation of how the system generates novelty in order to increase the range of alternatives available within the system and, thereby, foster diversity. The approach should also explain the process of selection from a range of diversity, where selection takes time and accommodates a number of alternatives concurrently, allowing different firms, processes and technologies to co-exist and compete. Finally, McElvey suggested adapting the approach by directly addressing the cause of an evolutionary pattern of change, which is to be found in the decisions and actions of various agents operating in an institutional or organisational context.71
As the above list of ‘building blocks’ suggests, Malerba placed knowledge at the foundation of technological change and accorded knowledge a central role in innovation. He also joined other evolutionary writers in pointing out that “knowledge is highly idiosyncratic at the firm level, does not diffuse automatically and freely among firms and is absorbed by firms through their differential abilities accumulated over time.”72 As indicated elsewhere, an understanding of firm behaviour, including differences in the knowledge they accumulate, is a necessary, but not a sufficient, explanandum in any analysis of military technological innovation in small Western democracies. Hence in amplifying and refining Malerba’s ‘knowledge building block’ for the purposes of this thesis, one area of development will entail clarifying what McKelvey called the mechanisms by which information is retained and transmitted.
Malerba drew attention to the importance of technologies in defining and shaping the nature, boundaries and organisation of sectors. In particular, according to Malerba, “Links, complementarities among technologies, artefacts ad activities play a major role in defining the real boundaries of sectoral systems.”73 Malerba’s perspective of the role of technology in sectoral innovation systems accords with Carlsson’s notion of technological systems discussed earlier. However Malerba’s emphasis on the links and complementarities among technologies is better suited to an investigation of how defence innovation fosters technological diversity. To put this point another way: Malerba’s ‘technology building block’ is sufficiently flexible to accommodate the tendency of military technological innovation to subsume the outputs of several technological systems.
The technological dimension of Malerba’s sectoral system of innovation is therefore well suited to analysis of military technological innovation. In order to illuminate the causal connections involved, however, Malerba’s ‘technology building block’ needs to be augmented by analysis of what McKelvey has referred to as the means by which technological novelty is generated so as to increase the range of alternatives available
71 Maureen McKelvey, Using evolutionary theory to define systems of innovation, in Edquist (ed.) Systems of Innovation, pp. 201-220. 72 Malerba, p. 19. 73 ibid. 26
within the system and, thereby, foster diversity. Malerba himself lay the foundation for such modification in his emphasis on the key role played by demand for the products which are the outputs of his sectoral systems of innovation.
In outlining the role of demand in his sectoral systems of innovation, Malerba eschewed any notion of demand as an aggregate set of similar buyers and insisted that, rather, demand consists of heterogeneous agents, whose interaction with producers is shaped by institutions.74 Malerba’s emphasis on the heterogeneity of demand-side actors is particularly useful in the present context where, as already indicated, the government of a nation state is a monopsonist buyer of bona fide defence goods and services. By recognising the heterogeneity of demand-side actors, Malerba’s sectoral systems of innovation concept can accommodate the contested nature of defence demand flowing from the competing interests within a nation state’s defence organisation within the nation more broadly. By extension, Malerba’s approach recognises the impact on innovation performance of nationally specific institutions developed to mediate such contests.
Malerba went on to argue that demand combines with technologies to define, firstly, the nature of the problems that firms have to solve in their innovative and productive activities and, secondly, the types of incentives and constraints imposed on behaviours and organisations. Recognition of demand as a variable greatly enhances the utility of the sectoral systems of innovation approach in analysing military technological innovation: Malerba’s reference to problem solving in this context accords with the emphasis placed by Carlsson and Granberg (see earlier discussion) on solving technological problems as the prime impetus to activity in technological systems. Particularly significant for the purposes of this thesis is the way in which the concept of demand for solutions to technological problems informs McKelvey’s focus on selection from a diverse range of technological candidates as a prime determinant of causal relationships among the variables accounting for military technological innovation.
In his discussion of the actors and networks building block of his sectoral systems of innovation, Malerba placed firms at centre stage. The pivotal role played by firms in the sectoral systems of innovation approach stems from their involvement in the innovation, production and sale of sectoral products and in the generation, adoption and use of new technologies.75 Particularly relevant for present purposes, however, is Malerba’s acknowledgement of the role played by non-firm organisations – for example, universities, financial organisations and government agencies – in the innovation process generally and, more particularly, in determining the characteristics of specific sectoral systems of innovation.
74 ibid., pp. 42-43. 75 ibid., pp. 24-26. 27
As the above discussion of the national systems of innovation approach suggests, this thesis seeks to illuminate the role of non-firm actors in military technological innovation. For example, the international diffusion of military technology is an important aspect of the military technological innovation phenomena that involves not only firms (which undertake most of the transactions involved) but also governments (which regulate the international arms trade). Similarly, the monopsonist characteristics of the bona fide market for military technology suggests that military users and other non-firm actors on the demand side also influence the outcomes of the military technological innovation process. Malerba’s sectoral systems of innovation approach provides a useful framework for bringing these non-firm, national-level actors into focus.
The utility of the sectoral systems of innovation approach in this regard is further enhanced by Malerba’s recognition of the importance of non-market as well as market relationships in connecting the heterogeneous agents that populate sectoral systems of innovation.76 As discussed earlier, military technological innovation takes place in an uncertain and changing environment. In this kind of environment networks emerge because they enable heterogeneous actors to access and integrate complementary sources of knowledge, capabilities and specialisation. As Malerba put it, “The types and structures of relationships and networks differ from one sectoral system to another, as a consequence of the features of the knowledge base, the relevant learning processes, the basic technologies, the characteristics of demand, the key links and dynamic complementarities.”77 Again, Malerba’s sectoral systems of innovation approach provides a useful framework for exploring how innovation outcomes are affected by the formal and informal relationships among the various firm and non-firm actors involved in the military technological innovation process.
Malerba echoed Edquist who defined institutions as the ‘rules of the game’, that is, as sets of common habits, norms, routines, established practices, rules or laws that regulate relations and interactions among individuals, groups and organisations.78 Malerba also echoed Metcalfe in noting that national institutions may have major effects on sectoral systems.79 In exploring the institutional dimensions of military technological innovation, this thesis builds on and extends this congruence of the national and sectoral systems of innovation approaches. In doing so, it takes into account McKelvey’s point that it is the decisions and actions of various agents operating in an institutional or organisational context that account for evolutionary patterns of change.
76 ibid., p. 25. 77 ibid., p. 26. 78 Edquist, Systems of innovation, in Fagerberg, Mowery and Wilson (eds) The Oxford Handbook of Innovation, p. 183. 79 Malerba, pp. 27-28. 28
2.5 Conclusion The literature on military technological innovation reviewed in this chapter does not explain why similar countries addressing comparable requirements for military capability differ in terms of the time they take to meet those requirements, the cost they incur in doing so and the pattern of diffusion/development the solutions to those requirements display following their acceptance. This suggests a gap in the military technological innovation literature. After reviewing the literature on innovation systems, the chapter concluded that Malerba’s sectoral systems of innovation model, suitably adapted, provides a satisfactory basis for addressing the above question and, by extension, the identified gap in the military technological innovation literature. The review also indicated, however, that Malerba’s sectoral systems of innovation model would need substantial modification before it can be used for this purpose. This adaptation is undertaken in Chapter 3, which is devoted to developing the notion of a defence sectoral innovation system as the theoretical framework for organising the case studies.
29
Chapter 3: Defence Sectoral Innovation Systems This chapter establishes the framework required to investigate why comparable nations with access to comparable technologies perform differently in generating innovative solutions to comparable military capability requirements. The framework so developed is called a defence sectoral innovation system and is based on Malerba’s generic sectoral systems of innovation model and his five ‘building blocks’ of knowledge, technology, demand, actors/networks and institutions. However these building blocks must be adapted to highlight those features that, firstly, distinguish defence sectoral innovation systems from other non-defence innovation systems and that, secondly, distinguish one defence innovation system from another. To this end Section 3.1 of the chapter places the building blocks in an order that is suited to the analysis of defence sectoral innovation systems and then defines their meaning and use in this thesis. In Section 3.2 each of the building blocks is discussed in more detail to provide a framework for analysis of the Swedish and Australian case studies. Section 3.3 summarises the questions and propositions concerning the structure and operation of a defence sectoral innovation system that emerged in Section 3.2 and that will be investigated in the case studies.
3.1 Defining and ordering the defence innovation building blocks This thesis places institutions at the beginning of its discussion of defence sectoral innovation systems. Malerba80 followed Edquist and Johnson in defining institutions as “sets of common habits, routines, established practices, rules, or laws that regulate the relations and interactions between individuals and groups”.81 McKelvey drew attention to the role of human agency in determining the purpose of institutions, their structure and the way they operate.82 This thesis modifies Malerba’s notion of institutions but retains McKelvey’s emphasis on the role of human agency in their structure and operation.
In this thesis, institutions are deep-seated, durable, culturally based and historically shaped ‘rules of the game’ or social and political constraints that condition innovation choices by national defence innovation actors. Institutions are deeply rooted in national identity and help define the boundary between a nation state and its external environment. Institutions also shape how the national defence innovation actors that populate a given defence sectoral innovation system interact with each other. Such institutions are a distinctive feature of defence innovation systems. They also set the parameters within which other elements of defence sectoral innovation systems interact in distinctive ways. The connection between institutions and the functioning of defence sectoral innovation systems are described in more detail in Section 3.2.
80 Malerba, pp. 27-28. 81 Charles Edquist and Bjorn Johnson, Institutions and organisations in systems of innovation in Charles Edquist (ed): Systems of Innovation, p 46. 82 McKelvey, pp. 208-209. 30
Institutions are connected to innovation activity by actors and networks. Accordingly, this thesis addresses actors and networks as the second building block in defence sectoral innovation systems. The thesis accepts Malerba’s proposition that actors respond to uncertain and changing circumstances by forming networks with actors with dissimilar, rather than similar, attributes. Such networks “allow access to and the integration of complementarities in knowledge, capabilities and specialisation.”83 However, this thesis extends Malerba’s conception of actors and networks in order to facilitate analysis of the role played by actors and networks in defence sectoral innovation systems.
The first extension of Malerba’s conception relates to the definition of ‘actors’. Malerba placed firms at the centre of his generic sectoral system of innovation. This thesis, however, follows Edquist and others in explicitly extending the range of actors involved in defence sectoral innovation systems.84 Accordingly, defence innovation actors range from, at one end of the spectrum, individuals (whose skills and knowledge drive military technological innovation) through groups (a number of people linked by some mutual or common relationship or purpose) to, at the other end of the spectrum, organisations. The latter are formal structures consciously constructed by humans to serve an explicit purpose. In this thesis, organisations are populated by individuals and groups directly or indirectly engaged in defence innovation activity and include both firms and related commercial entities as well as government agencies.
A second extension of Malerba’s conception of actors and networks relates to ‘networks’. This thesis follows Potts in according equal analytical significance to, on one hand, actors and, on the other hand, to the networks linking those actors.85 Organisations – including firms and government agencies – marshal factors of production and also establish the connections between those factors. In doing so actors at the individual, group and organisation levels of analysis do more than exchange information: they also make decisions and choices and act on the decisions and choices they make, thereby converting information into a form of knowledge.
Defence innovation actors use institutionally conditioned networks to exchange knowledge. Accordingly this thesis addresses knowledge as the third building block in the framework for analysing defence sectoral innovation systems. Malerba placed knowledge at the base of technological change and assigned knowledge a central role in his conception of sectoral innovation systems.86 Malerba’s conception of knowledge, however, is too broad and too general to provide the degree of fidelity necessary in distinguishing defence sectoral innovation systems from non-defence sectoral innovation systems and one defence sectoral innovation system from another. In order to meet the need for a more narrowly focused
83 Malerba, p. 26. 84 Charles Edquist and Bjorn Johnson, pp. 46-49 85 Potts, The New Evolutionary Microeconomics, especially pp. 59-60. 86 Malerba, p. 19. 31
and incisive treatment of knowledge in defence sectoral innovation systems, this thesis focuses on military doctrine. In Australian usage:
Military doctrine is an officially sanctioned, formalised and written expression of institutionally accepted principles and guidance about what armed forces do and how they do it. It contains fundamental principles by which military forces guide their actions in support of national objectives.87
In this sense, military doctrine is a sub-set of national security policy and grand strategy (both of which, in this thesis, are addressed in the institutions building block of defence sectoral innovation systems). The notion of military doctrine as a distinctive feature of knowledge in defence sectoral innovation systems is analysed further in Section 3.2. At this stage what matters is that, in order to achieve a military effect, military actors use more or less specialised artefacts (for example, rifles, submarines or combat aircraft). The design of such artefacts reflects, among other things, the military doctrine prevailing among those actors and the technology available to them.
Accordingly this thesis addresses technology as the fourth building block of a defence sectoral innovation system. This thesis defines technology as “tools or tool systems by which we transform parts of our environment, derived from human knowledge, to be used for human purposes.”88 At the most general level, the way humans use technology is conditioned by cultural norms and values and by social roles and practices. This thesis adopts a similarly inclusive notion of military technology in the sense that, to be considered successful, a military technological innovation must be embedded in a military socio- technical regime.
This inclusive notion of military technology helps define what constitutes a novel solution to a requirement for military capability. As the above discussion suggests, this thesis distinguishes between the technological content of an innovation and the embedding content of that innovation. The technological content of an innovation is its knowledge basis derived from science and practical wisdom. The embedding content of a technological innovation is about the arrangements needed to link the technological content into an operating system.89
In the military context, the technological content of an innovation is incorporated in the artefacts used to achieve a given military effect (for example, innovations in propulsion technology may increase the speed with which platforms can transport personnel and
87 D.J. Hurley., The Foundations of Australian Military Doctrine, Australian Defence Publishing Service, July 2012, Canberra, page 3.2, available at http://www.defence.gov.au/adfwc/Documents/DoctrineLibrary/ADDP/ADDP-D- FoundationsofAustralianMilitaryDoctrine.pdf accessed 9 July 2014. 88 Louis G. Tornatzky and Mitchell Fleischer , p. 10. 89 Donald C. Pelz, Fred C. Munson and Linda Jenstrom, Dimensions of Innovation, Journal of Technology Transfer, Vol 3 (1), 1978, pp. 36-37. 32
materiel). It is the technological content of a military technological innovation that defines that innovation’s military utility. The latter concerns the efficiency and effectiveness with which an artefact enables a user to achieve a military effect. Enhanced military utility is a necessary, but not a sufficient, condition for a military technological innovation to constitute a novel solution to a military capability requirement.
The embedding content of a military technological innovation encompasses the individual and organisational behaviours or processes that enable operators of the artefacts incorporating the technological content to use them to achieve a given military effect. The synthesis of artefact, knowledge and human behaviour constitutes a socio-technical regime.90 In the present context, a military socio-technical regime is manifest, firstly, in the formally constituted armed forces – for example armies, navies and air forces – maintained by nation states to protect and advance their respective strategic interests and, secondly, in the networks of actors – ranging from government organisations to companies – involved in supplying and supporting the specialist materiel each force uses to achieve a given military effect.
A key element of the embedding content of a military technological innovation is the operational knowledge of how to make effective use of an artefact in order to achieve a military effect. More generally, it is the embedding content of a military technological innovation that defines its military capability value. Enhanced capability value is both a necessary and a sufficient condition for a military technological innovation to constitute a novel solution to a military capability requirement. The military technology building block of defence sectoral innovation systems is analysed further in Section 3.2.
The above distinction between the military utility of an innovation and the capability value of that innovation leads to discussion of demand which constitutes the fifth and final building block of defence sectoral innovation systems. According to Malerba, “Together with technologies, demand defines the nature of the problems that firms have to solve in their innovative and production activities and the types of incentives and constraints on particular behaviours and organisations.”91 To help identify the causal linkages between demand and other elements of the defence sectoral innovation system, however, this thesis extends and focuses Malerba’s conception of demand. Such refinement begins with a recognition that defence is a public good, the demand for which is only partially determined by market signals.
In defence sectoral innovation systems, then, the primary chain of causation runs from, firstly, exogenous developments to, secondly, actors’ perceptions of those developments as causing a problem to, thirdly, recognition by actors that addressing the problem will require
90 Johan Schot and F. Geels, Niches in evolutionary theories of technical change: A critical survey of the literature, Journal of Evolutionary Economics, Vol 17, 2007, pp. 608-609. 91 Malerba, p. 29. 33
a new military capability to, fourthly, a decision by those actors to take action to acquire that new military capability. A secondary, weaker, chain of causation runs from recognition of a technological opportunity to recognition of an economic opportunity to a decision to exploit the economic opportunity by utilising the technological opportunity. Demand for military technological innovation is a product of the above primary chain of causation.
In defence sectoral innovation systems, then, the perception of a military problem is a necessary but not a sufficient condition for military technological innovation. Similarly, recognition that a military problem is sufficient to warrant new military capability is a necessary condition for military technological innovation but is not sufficient to initiate the innovation process. The necessary and sufficient conditions for military technological innovation are linked decisions by national actors to search for a solution to an identified requirement for a new military capability, then to select a solution to that requirement and, finally, to divert finite resources from other priorities to obtain that solution. Accordingly, in this thesis demand derives from the requirement for military capability. But the demand building block of defence sectoral innovation systems is defined in terms of the action taken, firstly, to search for one or more possible solutions to that requirement, secondly, to select a solution from the menu of candidate solutions so identified and, thirdly, to procure the solution so identified. The demand building block of defence sectoral innovation systems is analysed further in Section 3.2.
3.2 Identifying the distinctive features of defence innovation building blocks In this section each of the five building blocks comprising a generic defence sectoral innovation system is interpreted and extended. The amplified building blocks provide the framework used to organise, present and analyse the case studies of the Swedish and Australian defence sectoral innovation systems presented in Chapters 4 to 9. In the following paragraphs each of the building blocks is discussed with a view to facilitating their use in, firstly, identification of the distinctive features of the Swedish and Australian defence sectoral innovation systems and in, secondly, analysis of how those features affected innovation performance in the Swedish and Australian systems respectively.
3.2.1 Institutions Institutions, defined as deep-seated ‘rules of the game’, constitute the first of the five building blocks that comprise the model of defence sectoral innovation system used in this thesis. In military technological innovation, such institutions work to shape innovation choices, to encourage innovation in certain directions and, often, to constrain innovation choices in other ways. The following paragraphs describe three features of the institution building block that will be used in analysing the Swedish and Australian defence innovation systems in Part Two of the thesis. The first distinctive feature of the defence institution
34
building block is the cluster of objectives, perceptions and choices generally called national security policy and/or grand strategy.
This cluster of policy and strategy is nation-specific. It provides the ‘lens’ through which national defence actors perceive developments exogenous to that nation’s defence sectoral innovation system, influences the way those actors define national security interests affected by such developments and, by extension, influences the action they take to protect and advance those interests. Policy and strategy work through this chain of causation to influence judgements about the requirement for military capability. To the extent such judgements prompt demand for novel solutions to those requirements, they drive activity in the defence sectoral innovation system. For example, a key objective of the Swedish case study is to analyse how the institution of armed neutrality influenced Swedish military technological innovation. Similarly, the Australian case study will address the role of strategies like alliance-based collective security in shaping Australian innovation choices.
The second distinctive feature of the defence institution building block is the narrative or rationale used by defence actors in democracies to legitimise military technological innovation. Because national defence is a public good, and because investments in defence capability entail substantial opportunity costs, national defence actors in democracies need to articulate a generally accepted rationale for the nature and scale of the investments they make. To be persuasive, this rationale must enable the actors responsible to justify both the amount of domestic resources allocated to, and the way those resources are allocated to, generating and sustaining the military capability required to implement national security policy and/or execute grand strategy. Any such rationale is nation-specific and, to that extent, influences activity in that nation’s defence sectoral innovation system in distinctive ways. The rationale influences military technological innovation outcomes by legitimising, or constraining, the choices actors make in meeting requirements for military capability. In the Swedish case study, for example, the rationale for corporatist business arrangements is explained and the impact of these arrangements on Swedish innovation outcomes is assessed. Similarly, the Australian case study addresses the rationale of defence self- reliance and assesses the impact of that rationale on Australian innovation choices.
Governance constitutes the third distinctive feature of the defence institution building block. In this thesis governance denotes the generally accepted arrangements by which accountable public sector actors make choices in the course of developing and implementing national security policy and/or grand strategy. Governance is about how actors make military capability choices, as distinct from what military capability choices they make. The legitimacy92 of choices made by such actors in executing national security policy and/or grand strategy depends importantly on their complying with, and on their being seen to comply with, generally accepted and understood governance arrangements. For example,
92 For fuller discussion of the importance of legitimacy in innovation see Bergek, pp. 416-417. 35
the Swedish case study will address the influence on Swedish military technological innovation of the autonomy enjoyed by Swedish government agencies, while the Australian case study will consider the influence on Australian military technological innovation of open and effective competition for defence business.
3.2.2 Actors and networks Actors and actor networks constitute the second of the five building blocks that comprise the model of defence sectoral innovation system used in this thesis. The term ‘actor’ encompasses individuals, groups and organisations who make decisions in the course of their participation in the process of military technological innovation and act on the choices so made. Such choices may range from ranking requirements to choosing between ‘make’ or ‘buy’ solutions to selecting suppliers of those solutions. The characteristic that distinguishes actors in defence sectoral innovation systems from those in other such systems is the competence with which they make the choices and perform the tasks involved in military technological innovation. The term ‘network’ encompasses the links between actors with complementary competencies that allow those actors to exchange the knowledge required to design, develop and produce solutions to requirements for military capability. Networks linking such actors in one defence innovation system can be distinguished from the networks in other defence innovation systems (and from networks in non-defence innovation systems) by the efficiency and effectiveness with which they enable specialist actors to marshal and deploy the diverse competencies required to generate novel solutions to military capability requirements. This entails those actors exchanging information about military capability requirements and novel solutions to those requirements.
These two distinguishing characteristics of defence innovation actors and networks need to be analysed in greater detail. The following analysis addresses three themes. The first theme concerns the notion of competence, how it is acquired and by whom. The second theme concerns the links between, on one hand, the notion of competence and, on the other hand, the discrete functions performed by the actors populating defence sectoral innovation systems. The third theme concerns the networks that enable actors with particular competencies performing particular functions to link to other actors with complementary competences performing complementary functions in the innovation system.
According to Foss:
Competence is a typically idiosyncratic knowledge capital that allows its holders to perform activities – in particular solve problems – in certain ways and typically to do this
36
more efficiently than others. Because of its skill-like character, competence has a large tacit component, and is asymmetrically distributed.93
Competence results from the specific combination of knowledge, skills and resources of particular actors. Defence-specific competencies distinguish defence actors from non- defence actors. In defence sectoral innovation systems, skills are manifest at the level of the individual human actor but competence is manifest at the level of the team, group and organisational actor. Actors in defence sectoral innovation systems do not bring such competencies to the defence innovation task fully formed. Rather, they develop those competencies through learning. This thesis follows Foray in distinguishing between learning offline and learning online. Learning offline involves learning by formal R&D. Learning online involves learning in the course of production (learning-by-doing), learning in the course of using an artefact or service (learning-by-using) and learning through the interaction of user and producer.94
In defence sectoral innovation systems, learning offline through R&D is undertaken at some distance from the process of production (and utilisation). This distance can be spatial, temporal and institutional. The efficiency and effectiveness with which actors in defence sectoral innovation systems can generate novel solutions to military capability requirements depends on the efficiency and effectiveness with which actors can bridge the distance between such learning offline and the process of production/utilisation. In defence sectoral innovation systems, the efficiency and effectiveness with which actors build bridges between R&D and production/utilisation is influenced by, among other things, the efficiency and effectiveness with which they experiment. Such experimentation is necessary to manage the uncertainty which, despite the digital revolution,95 continues to characterise military technological R&D.96 Such uncertainty means that in the course of such experimentation, a military technological development may ‘fail’ to demonstrate the requisite degree of military utility but may, nevertheless, generate valuable information. One defence sectoral innovation system will differ from another in how, and the extent to which, the actors involved learn from such ‘failures’.
Defence innovation systems that treat experimental ‘failures’ as the inevitable but unknown costs of experimentation (what Potts termed ‘good waste’) will tend to accumulate knowledge. Such accumulation reduces the gap between the extant stock of knowledge in the system and that required to generate novel solutions to military capability
93 N. Foss, The emerging competence perspective in N. Foss and C. Knudsen (eds): Towards a Competence Theory of the Firm, London, Routledge, 1996, p. 1. 94 Dominique Foray, The Economics of Knowledge, MIT Press, Cambridge, 2004, pp. 50-54. 95 Henry Ergas, The ‘new face’ of technological change and some of its consequences, unpublished memorandum, March 1994, cited by Paul David and Dominique Foray, Accessing and expanding the science and technology base, Science Technology and Industry Review, OECD, No 16, Paris, 1995, p. 43. 96 See Arthur J. Alexander, The Linkage Between Technology, Doctrine, and Weapons Innovation: Experimentation for Use, The RAND Corporation, Santa Monica, May 1981, especially pp. 17-22. 37
requirements. Conversely, innovation systems that encourage actors to abandon the experiment and discard the associated learning (what Potts called ‘bad waste’) incur extra costs stemming from a failure to harvest the results of experimentation in terms of learning.97 Such systems are inefficient in the sense that they will tend to dissipate knowledge, thereby widening the gap between the extant stock of knowledge and that required to generate novel solutions to military capability requirements. The Swedish and Australian case studies draw attention to the impact on innovation outcomes of differences in how, and the extent to which, Swedish and Australian actors managed the experimentation inherent in military technological innovation.
In Foray’s usage, learning online occurs either in the course of production or in the use of an artefact or service. This thesis adopts Foray’s approach to online learning. Such online learning includes learning-by-doing and learning-by-using. The most rudimentary form of learning-by-doing is essentially about the incremental development of expertise as a result of repeating a task and does not entail technical or organisational change.98 In defence sectoral innovation systems such rudimentary learning-by-doing is manifest at the level of individual skill and is less apparent in organisational competencies. Hence such rudimentary learning-by-doing is not a distinctive feature of defence sectoral innovation systems.
The next level of learning-by-doing identified by Foray is more explicitly ‘cognitive’ in the sense that it entails undertaking experiments within the constraints imposed by the need to continue production of the goods and services involved. Such constrained experimentation is aimed at improving production processes, artefact design or service delivery:
Through these experiments new options are spawned and variety emerges. This is learning based on an experimental concept, where data is collected so that the best strategy for future activities can be selected. Technical and organisational changes are then introduced as a consequence of learning by doing. The locus of the learning process is not the R&D lab but the manufacturing plant or site of use.99
Defence sectoral innovation systems differ widely in terms of how, and the extent to which, the actors involved capture such learning-by-doing. One differentiator is the opportunity and incentive for actors to apply in a later production activity the knowledge they have gained through learning-by-doing in an earlier activity. A second differentiator is the ‘gap’ between the stock of knowledge accumulated by actors through learning-by-doing in the course of producing a novel solution to one requirement and that knowledge required to produce a novel solution to a subsequent, more demanding requirement. In general, the more military requirements evolve in a path-dependent way, the smaller is the gap between
97 Jason Potts, The innovation deficit in public services: the curious problem of too much efficiency and not enough waste and failure, Innovation: Management, Policy and Practice, Vol 11 (1), April 2009, pp. 34-43. 98 Foray, p. 61. 99 ibid. 38
the extant stock of knowledge and the knowledge required to meet the evolved requirement. By extension, path-dependent evolution of requirements will tend to enhance the relevance of prior learning-by-doing in solving that requirement. The Swedish and Australian case studies will explore how innovation outcomes were affected by differences in opportunities and incentives for the actors in each system to learn-by-doing.
Defence sectoral innovation systems also differ in how, and the extent to which, the actors involved capture learning-by-using. Such learning generally occurs when:
Faced with new and unexpected local situations, users have to solve problems that designers failed to anticipate, and are thus in a position to teach and inform those who design systems. It is because there are limits to the perfect reproduction of the environment during the R&D phases that problems arise in the course of normal production and use.100
Such learning-by-using becomes a particularly important differentiator among those defence innovation systems that design, develop, produce and operate military systems characterised by a high degree of systemic complexity. In general, military operators of complex systems will accumulate very substantial tacit knowledge regarding the operational characteristics of the system, its true potential and how these relate to design and other features of the system.101 The Swedish and Australian case studies address the implications for innovation outcomes of differences in how, and the extent to which, actors in each system fed such learning-by-using back to the system designers and how those designers took that learning-by-using into account in devising solutions to the next generation of requirements.
This raises the functions performed by the actors populating defence sectoral innovation systems – the second theme of this sub-section. The following paragraphs identify different actor categories and then investigate how these different actor categories influence defence innovation outcomes in distinctive ways. This requires an extension and refinement of Malerba’s much more generic conception of actors and networks. To this end, this thesis uses Eliasson’s competence bloc theory.102 In Eliasson’s usage, a competence bloc comprises actors with the various competencies needed to generate, identify, select, develop and exploit new business ideas successfully. Such a bloc works through the dynamic interaction of people or groups of people embodying the tacit competencies required to perform these functions – what Eliasson called ‘dynamic functionality’.103
100 ibid., pp. 62-64. 101 For a fuller treatment of this proposition see Nathan Rosenberg, Inside the Black Box: Technology and Economics, Cambridge University Press, Cambridge, 1982, pp. 122-123. 102 Gunnar Eliasson, Competence blocs in the experimentally organized economy in Gunnar Eliasson (ed.), The Birth, the Life and the Death of Firms: The Role of Entrepreneurship, Creative Destruction and Conservative Institutions in a Growing and Experimentally Organized Economy, The Ratio Institute, Stockholm, 2005. 103 Ibid., p. 56. 39
Eliasson’s competence bloc theory explicitly recognises the importance of tacit knowledge in the innovation process. It takes into account the costs inherent in generating and communicating information and, in particular, recognises that converting tacit knowledge into communicable information is expensive and may entail prohibitive loss of content. The theory also treats business mistakes and the failure of experiments as the practical consequence of bounded rationality and a normal cost of innovative development.104 As explained in the following paragraphs, the dynamic functionality of Eliasson’s competence bloc depends on mutually supportive links among customers, innovators, entrepreneurs, venture capitalists and industrialists.
Eliasson accorded the informed customer a pivotal role in the innovation process. Such customers value an innovation. Their willingness and ability to pay for that value sustains the innovation process. Hence the Swedish and Australian case studies identify, and analyse the role played by, actors performing the defence customer function. The case studies are informed by Eliasson’s argument to the effect that the more advanced and radically new the product and technologies involved, the more important the customer becomes to the innovation process:
In the long term … the quality of the products will be limited from above by the quality of customers’ understanding of the usefulness of the product, their willingness to pay and their contribution of user knowledge to the development of the new product/technology.105
In Eliasson’s model, innovators combine old and new technologies into new, composite technologies. The Swedish and Australian case studies identify, and analyse the role played by, actors performing the innovator function. The case studies are informed by Eliasson’s argument that innovation outcomes are driven by the way innovators combine technologies (both old and new) in novel ways or apply them to solve new problems. The case studies also take into account Eliasson’s proposition that the supply of novel technological combinations is a necessary but not a sufficient condition for innovation:
Technologies have to be identified and commercialised to result in economically and socially valuable output ... and this is the phase when critical project selection by economic criteria occurs, large resources have to be mobilised and business mistakes are committed.106
In Eliasson’s model it is entrepreneurs who have the skill required to pick, ex ante, what innovative combination of technologies will be profitable. The Swedish and Australian case
104 Ibid., pp. 27-28. 105 Gunnar Eliasson, Advanced Public Procurement as Industrial Policy: The Aircraft Industry as Technical University, Springer, New York, 2010, p. 43. 106 ibid., p. 44. 40
studies identify, and analyse the role played by, actors performing the entrepreneur function. The case studies are informed by Eliasson’s argument that:
The task of the entrepreneur is to identify commercial winners among the suppliers of innovations and to get his/her technology choice onto a commercial footing. The understanding of the entrepreneur may be of a long run nature, or more temporary in the sense that they may have to reconsider their decision or make a business mistake. The main thing is that the entrepreneur acts on the perceived opportunity ...107
Eliasson’s competence bloc makes explicit provision for venture capitalists who not only provide early finance for start-ups but who are also industrially competent selectors of entrepreneurs. The Swedish and Australian case studies identify, and analyse the role played by, actors performing the venture capitalist function. The case studies are informed by Eliasson’s argument that venture capitalists understand the entrepreneurial choice and, on that basis, provide the risk capital that entrepreneurs require to commercialise the innovation they select. Subject to the quality of assessment and judgement involved, venture capitalists reduce the incidence of business mistakes. Venture capitalists are rewarded for the quality of their judgement in the form of capital gains on their equity in the entrepreneur’s venture.
Industrialists constitute the final element of Eliasson’s competence bloc. It is industrialists who have the functional competence required to carry the selected winner on to industrial scale production, marketing and distribution. The Swedish and Australian case studies identify, and analyse the role played by, actors performing the industrialist function. The case studies are informed by Eliasson’s argument that it is the existence of an industrialist willing and able to take the entrepreneur’s artefact to the market that is the keystone of the structure of incentives required for the functioning of the competence bloc as a whole.
The third theme of this discussion of the actors and networks building block concerned the links – networks – among actors with complementary competences. The performance of a competence bloc depends on the connections among the actors performing the customer, innovator, entrepreneur, venture capitalist and industrialist functions. Deficiencies in the competence of, or the absence of, any one of these actor groups tends to undermine the incentive structure that is critical to the bloc’s functionality. Similarly, deficient or absent linkages among actors performing the competence bloc functions will cause comparable innovations to take longer, cost more and diffuse less than would otherwise have been the case. Hence the Swedish and Australian case studies examine the connections between and among the customers, innovators, entrepreneurs, venture capitalists and industrialists populating each system.
107 Eliasson, Competence blocs, in Eliasson (ed) p. 60. 41
3.2.3 Defence-specific knowledge The third building block of defence sectoral innovation systems relates to defence-specific knowledge. Much of the defence-specific knowledge that directly affects military technological innovation has already been mentioned in the institutions building block discussed above or will be addressed in the technology building block discussed below. Military doctrine constitutes a third body of defence-specific knowledge that shapes choices by actors translating requirements for military capability into novel solutions to those requirements. Military doctrine is about, firstly, what military means actors employ to achieve the objectives of national security policy and grand strategy and, secondly, how those actors employ such military means to achieve those objectives.108
In his classic articulation of the connection between military doctrine and military technological innovation, Holley argued that superiority in weapons stems not only from selecting the best ideas from advancing technology but from relating those ideas to a doctrine or concept for their tactical or strategic application.109 Holley analysed the US Army’s failure in World War One to exploit the emergent aircraft technology as a failure not only of developing and deploying aircraft suitable for that conflict but also of doctrinal neglect. The latter meant that the US Army’s Air Arm failed to husband a body of experience from which it could derive an acceptable concept of air warfare. Because it lacked such a concept or doctrine, the Air Arm had little basis for providing authoritative direction to the development of aircraft for the future.110
Alic draws attention to the double-edged nature of military doctrine: military organisations formulate doctrine – warfighting principles and practices often set down in written manuals – to guide the exercise of discretion in battle and as an antidote to confusion and fear. But those same doctrines can become part of a belief system that stifles fresh thinking and inhibits innovation.111 Military doctrine is also informed by the lessons that actors – whether individuals, groups and formal organisations – learn over time about the use of military forces in support of national policy. As Bayerchen has shown, it incorporates individual and collective experience as well as lessons learned from the experience of allies, adversaries and like-minded forces.112 Finkel showed how such learning causes military doctrine to co- evolve with military operations and tactics.113 Posen showed how military doctrine is
108 B.R. Posen, The Sources of Military Doctrine – France, Britain and Germany Between the World Wars, Cornell University Press, Ithaca and London, 1984, pp. 13-14. 109 I.B. Holley, Ideas and Weapons, Archon Books, Hamden, Connecticut, 1971, p. 14. 110 ibid., p. 176. 111 Alic, p. 16. 112 A. Bayerchen, From radio to radar: interwar military adaptation to technological change in Germany, the United Kingdom, and the United States in W. Murray and A. Millett (eds), Military Innovation in the Interwar Period, Cambridge University Press, Cambridge, 1996, especially pp. 269-275. 113 M. Finkel, On Flexibility: Recovery from Technological and Doctrinal Surprise on the Battlefield, Stanford University Press, Stanford, California, 2011, especially pp. 223-231. 42
influenced by strategic geography and, to a lesser extent, by technology and how it evolves in response to political circumstances, national security policy and national strategy.114
The Swedish case study explores how Sweden’s distinctive ground-based air defence doctrine influenced the way the actors in the Swedish defence sectoral innovation system generated a novel solution to the SwAF requirement for early warning of an airborne assault. Similarly, the Australian case study analyses how Australia’s notion of core force influenced the Australian search for and development of a system for surveillance of the continent’s northern maritime approaches.
3.2.4 Technology Technology constitutes the fourth of the five building blocks that comprise the model of defence sectoral innovation system used in this thesis. In order to facilitate identification of the causal links between technology and defence innovation outcomes, the discussion of technology begins with an analysis of technological paradigms. This is followed by analyses of the technology systems underpinning those paradigms and of the trajectories along which those systems develop over time. The discussion concludes with a description of the main technological change drivers that account for the development of technological systems along those trajectories over time.
A technological paradigm is “a pattern for solutions of selected techno-economic problems based on highly selected principles derived from the natural sciences.”115 By extension, a military technological paradigm is a pattern for solution of selected requirements for military capability based on the technological knowledge required to design and produce military artefacts, the operational knowledge required to make effective use of those artefacts and the user knowledge of the performance characteristics of an artefact.
Actors in defence sectoral innovation systems adopt a military technological paradigm which then influences the way they define the relevant problems to be addressed and what pattern of enquiry they will pursue (including the scientific principles to be applied and the material technology to be used). A military technological paradigm embodied in a system, platform or system-of-systems is also an exemplar of an artefact with a particular set of techno-economic characteristics that is to be developed or improved. That same military technological paradigm is also a set of heuristics that helps actors to decide next steps, to determine where and how to search, and to select a body of knowledge to draw on in doing so.116
114 Posen, pp. 239-241. 115 G. Dosi, The nature of the innovation process in G. Dosi, C. Freeman, R. Nelson, G. Silverberg and L. Soete (eds) Technical Change and Economic Theory, Pinter, London, 1988, p. 224. 116 G. Dosi, Sources, procedures, and microeconomic effects of innovation, Journal of Economic Literature, Vol 26, September 1988, p. 1127. 43
A military technological paradigm subsumes the hardware and software components of military technology discussed earlier. In defence sectoral innovation systems, a military technological paradigm results from the fusion of scientific, technological and operational knowledge into the successful solution of a strategic or operational problem. A military technological paradigm grows out of the trials and errors of individuals, firms and military actors at varying levels of aggregation. In addition, and as Eliason and Goldman argued, defence actors have a strong incentive to emulate military technological paradigms that have demonstrated superior performance on the battlefield, leading to the diffusion of those paradigms within and among defence sectoral innovation systems.117
In competing for military advantage, national defence actors will invest in a portfolio of military capabilities. Because defence is a public good undertaken by governments of nation-states, those actors inhabit a military socio-technical regime with distinctive national characteristics. The actors make idiosyncratic investments which reflect the nation-specific problems confronting those actors, the information they have about successful and unsuccessful solutions to comparable problems elsewhere, the menu of technologies from which they can choose and the nature and scale of the resources available for allocation to the capability development task. The Swedish and Australian case studies illustrate how such investments are influenced by the way the prevailing military technological paradigm conditions how relevant actors view the problem to be solved, frame the requirement for military capabilities to solve it and exercise demand for solutions to that requirement.
By influencing actors in this way, military technological paradigms will drive the search for, and influence the design, development and procurement of, the artefacts used by national defence actors to perform a specific military function. To the extent those artefacts have the functionality required to create the military effect desired by defence actors, their use will reinforce the prevailing military technological paradigm.
In defence innovation systems the primary chain of causation flows from nation-specific strategic perceptions through organisation-specific military doctrine through actor-specific military technological paradigms to consequential change in national military capability portfolios. If defence actors judge that, by extrapolating the prevailing military technology paradigm they will continue to achieve satisfactory outcomes in the competition for military advantage, then they will have an incentive to invest in path-dependent innovation within the framework of that paradigm. This incentive will be reinforced by the cost defence actors must incur in embedding an innovation in an established socio-technical system. The more novel the technological content of the innovation, the larger the adjustment of the prevailing socio-technical system required to use it to best effect, the higher the cost involved and the greater the disincentive to adopt the innovation.
117 Leslie C. Eliason and Emily O. Goldman, Theoretical and comparative perspectives on innovation and diffusion in Emily O. Goldman and Leslie C. Eliason (eds) The Diffusion of Military Technology and Ideas, Stanford University Press, Stanford, California, 2003, pp. 1-30. 44
Conversely, if defence actors judge that an extrapolation of prevailing military technology paradigms will not yield satisfactory outcomes in the competition for military advantage, then they are likely to be more receptive to disruptive innovations that lead to new military technology paradigms. Specifically, defence actors will have a strong incentive to emulate a military technological paradigm that has demonstrated superior performance on the battlefield and to incur the adjustment costs inherent in embedding that paradigm in their military socio-technical regimes.
The military technological paradigm that prevails among the actors populating a given military socio-technical regime at a given point in time does not emerge fully formed. Nor are such paradigms homogenous. Rather, they are the evolutionary result of at least three intertwined learning processes: some actors learn by using artefacts to perform a given military function and to achieve a given military effect. Other actors learn by designing, developing and producing those artefacts and their replacements at different levels of aggregation in the military technological hierarchy. Yet other actors learn by the interaction of both producers and users. The Swedish and Australian case studies show how differences in such learning caused innovation performance to diverge.
This evolutionary dynamism precludes homogeneity: the diversity of strategic knowledge among nation-states, the plurality of actors involved, the variety of military functions to be performed and the range of items encompassed in the military technological hierarchy mean that a military technological paradigm will normally encompass numerous specific technologies. The notion of technological systems provides a useful lens through which to explore the link between, on one hand, the evolution of military technological paradigms and, on the other hand, change in the heterogeneous technological bundles they subsume.
Technological systems comprise “knowledge and competence networks supporting the development, diffusion and utilisation of technology in established or emerging fields of economic activity.”118 National defence actors operate military artefacts that subsume numerous, co-evolving technological systems. National defence actors will assemble those artefacts to produce nation-specific portfolios of military capabilities tailored to solve nation-specific military problems. As the earlier discussion of military technology also indicated, technological systems may be adapted for military purposes but extend beyond military applications and into the wider technology base underpinning the economy as a whole. By identifying and searching technological systems, defence actors identify non- military technologies that may be used in solving a requirement for military capability. Conversely, it is through technological systems that military innovations diffuse into a wider non-military technology base.
118 Carlsson, p. 2. 45
Technological systems are dynamic, not static and immutable. Within each system the configuration of actors and institutions will change over time. This dynamism precludes any actor in a technological system identifying all possibilities. If actors’ knowledge and information processing capability is limited, then the competence of technological actors will be differentiated but in each case will be fairly stable and path-dependent, with the search for new knowledge being local.119
The bounded rationality of defence innovation actors means that defence technological paradigms and, by extension, defence sectoral innovation systems, are emergent and not optimal. Finally, the bounded rationality of defence innovation actors also helps account for what Dosi called the “stickyness” of the basic technical imperatives, rules of search and input combinations that characterise each technological paradigm.120 The Swedish and Australian case studies show how technological paradigms, once established among the actors populating a given socio-technical regime, help explain the durability of national characteristics of defence sectoral innovation systems despite the diffusion of military technological knowledge internationally.
Technological trajectories result from technological choices made by boundedly rational actors who operate within a given technological paradigm while orchestrating inputs from technological systems in order to address problems or exploit opportunities. A technological trajectory is the activity of technological progress along the economic and technological trade-offs defined by a technological paradigm.121 To paraphrase Dosi, the rate and direction of such technological progress derives from a combination of specific incentives stemming from the problems and opportunities perceived by actors populating a given socio-technical regime and the paradigm-bound, cumulative and local nature of technological learning by those actors. The resulting trajectories are defined by actor-specific sets of knowledge and expertise which combine with specific infrastructures and institutions associated with the generation and/or exploitation of specific skills to render progress along the trajectories essentially irreversible.122
The selection process is a key determinant of the development of technological trajectories in the military context. Military technological trajectories emerge when the selection process preserves favourable variants and gradually builds fitness in a direction determined by the nature of the selection environment fostered by markets, users and institutional factors. The impetus for movement along a given technological trajectory is provided by the demand for improved military utility at the component, sub-system and system levels of the defence technological base. Movement along a given technological trajectory is prompted
119 ibid., pp. 4-5. 120 Dosi, Sources, procedures, and microeconomic effects of innovation, p. 1142. 121 Carlsson, p. 225. 122 Dosi, Sources, procedures, and microeconomic effects of innovation, p. 1143. 46
by what Mokyr called micro-innovations.123 These are small, cumulative variations resulting from price signals and other institutional incentives. In the military context, such variations are the product of learning-by-doing and learning-by-using.
At the technological system level, this pattern of path-dependent development along a technological trajectory via a sequence of micro-innovations is a form of dynamic stability. The dynamic stability of military technological systems is entirely compatible with adaptive change via small but continuous and cumulative technological adjustments which, over time, can lead to new military technological paradigms. However, military technological innovation is also occasionally punctuated by what Mokyr called macro-inventions. These are developments – like radar – that lack a clear-cut parentage, constitute a clear break from previous technique and do not fit into the prevailing socio-technical regime.124 Such macro-inventions often first appear as ‘hopeful monstrosities’ which promise new technical and functional possibilities despite initially low performance that renders them uncompetitive with existing, more efficient technologies.
In defence sectoral innovation systems, existing military technologies are often embedded in a set of complementary technologies and nurtured by actors with a substantial investment in the prevailing military technological paradigm. These factors all work to raise the cost of the adjustment of the prevailing socio-technical regime required to embed the emergent ‘hopeful monstrosities’. In these circumstances, macro-inventions are rare events which, in order to develop sufficiently to withstand competition from established alternatives and to overcome the indifference of those content with the status quo, typically require dramatic exogenous changes in the external environment within which the relevant defence sectoral innovation system operates. The Swedish and Australian case studies show how entrepreneurs first fostered niche development of broad area surveillance technologies and then secured acceptance of those technologies as part of a portfolio of military capabilities.
Both micro-inventions and macro-inventions are the product of technological change drivers which foster new technological paradigms, prompt the development of technological systems and drive the evolution of technological trajectories. Learning by defence innovation actors is a fundamental change driver. Additional, more circumscribed technological change agents include, firstly, the design of military artefacts; secondly technological speciation; thirdly, niche development; and fourthly the diffusion of military technology.
This thesis follows Simon in placing design – conceived as any course of action aimed at changing existing situations into preferred states – at the core of all professional training. In
123 J. Mokyr, The Lever of Riches, Oxford University Press, New York, 1990, p. 291. 124 ibid. 47
knowledge terms, it is design activity that distinguishes the professions from the sciences.125 Alic extended Simon’s logic in describing engineering design as management of trade-offs and compromise among competing objectives. Each choice has consequences, feeding forward to shape and constrain future choices and in some cases also feeding back to force reconsideration of earlier decisions. The final configuration of an artefact will reflect trade- offs at both the system level and in the details of sub-systems and components.126 The design of military artefacts is an engineering activity that connects, on one hand, actors in the customer element of the defence competence bloc to, on the other hand, those actors performing the innovator, entrepreneur and industrialist functions of that competence bloc. The Swedish and Australian case studies show how design activity influences innovation performance through the efficiency and effectiveness with which it applies technological learning to the development of artefacts geared to the solution of a requirement for military capability.
Design is a knowledge-intensive activity that encompasses both knowledge integrating processes and knowledge brokering processes.127 The role of design as a knowledge integrating process is particularly apparent in the development of novel components, sub- systems, systems and platforms for military purposes. Such development entails the orchestration of inputs from many different areas of technical expertise. The Swedish and Australian case studies show how innovation performance is affected by the efficiency and effectiveness with which designers performed a knowledge integrating function by mediating the integration of diverse knowledge held by actors populating distinctive technological systems.
The role of design as a knowledge brokering process is particularly apparent in the process of embedding an artefact into a socio-technical system to meet a requirement for military capability. Here military users will need to adapt pre-existing operational knowledge and to gain sufficient mastery of the new artefact to prevail in the competition for military advantage. In designing the requisite adjustments to the relevant socio-technical regime, users will broker knowledge about the problem to be addressed, about the capability required to address it, about the characteristics of the artefact, and about the competencies required to use the artefact to generate the capability required to address the problem. Such knowledge brokering will be strongly influenced by the distinctive military doctrines discussed earlier. The Swedish and Australian case studies show how innovation performance is affected by the efficiency and effectiveness with which designers performed a knowledge brokering function by mediating the integration of diverse knowledge held by actors populating distinctive technological systems.
125 Herbert Simon, The Sciences of the Artificial (Second Edition), MIT Press, Cambridge, 1981, pp. 132-133. 126 Alic, pp. 116-119. 127 P. Bertola and J. Teixeira, Design as a knowledge agent: how design as a knowledge process is embedded into organisations to foster innovation, Design Studies Vol 24 (2), March 2003, pp. 181-194. 48
In military technological innovation, design competencies are the prime vehicle by which the ideas and abstract concepts contributed by numerous actors with diverse areas of expertise are defined, negotiated and codified in the process of producing a tangible, material artefact. The design process creates ‘models’ of artefacts intended to meet a requirement for military capability. By mediating the discussion of abstract concepts like military capability requirements and technological systems, the design process helps build shared agreements among stakeholders about the attributes of new artefacts and the trade-offs involved.128
There is nothing deterministic about the process of designing complex military artefacts to provide novel solutions to requirements for military capability. The process of design and development is typically iterative, involves considerable experimentation and validation, all of which involves learning and takes time. During this process of design and development, novel solutions are acutely vulnerable, particularly if they constitute a challenge to the established technological paradigm.
In order to protect novel solutions during their vulnerable design and development stage, the innovators and entrepreneurs involved will typically adopt a niche strategy. A niche can be defined in terms of the conditions which enable one or more producer actors (often, but not always, firms) to mobilise the resources needed to meet a particular segment of demand in a way that other actors cannot or do not.129 Niches develop as the product of human agency. They are the seedbeds for new technological paradigms, technological systems and technological trajectories. The extent to which niches are isolated from the prevailing socio-technical regime varies. In defence sectoral innovation systems, however, the isolation of military socio-technical regimes may result from spatial or geophysical factors, cognitive distance, socio-cultural factors and policy choice.130
Niches can affect military innovation outcomes despite the absence of a market if actors or constituencies are prepared to invest time and money in nurturing and developing a fledgling, non-profitable innovation. Subject to the prevailing institutional environment, actors and constituencies will make such investments because they judge that the innovation will become viable in the future, either through technical improvements or changing selection criteria, and thereby help realise highly valued social or collective aims. In the military context, niche-based learning processes are not only about technology but also about the articulation of user preferences and required changes in the prevailing military socio-technical regime. Selection thus becomes a co-evolutionary process involving
128 ibid., pp. 185-186. 129 Peter Hall and Robert Wylie, Isolation and technological innovation, Journal of Evolutionary Economics, Vol 24, 2014, especially pp. 359-361. 130 Johan Schot and F. Geels, Niches in evolutionary theories of technical change: a critical survey of the literature, Journal of Evolutionary Economics, Vol 17, 2007, pp. 616-619.
49
the perceived problem, the emergent technology and the prevailing military socio-technical regime.
Finally, and when a military innovation emerges in an environment lacking clear selection criteria, niches can provide ‘proto-markets’ for the new technology. In defence sectoral innovation systems, such proto-markets enable users and producers to interact in protected spaces. The nurturing and quasi-evolutionary learning processes may eventually result in the articulation of a clear demand, in turn paving the way for a technological speciation event. The latter occurs when an existing technology is used – often with little or no change – in a new domain of application, resulting in a disruptive innovation that cuts across prevailing paradigms and diverges from established trajectories.131
This thesis follows Levinthal in locating the mechanism for technological speciation in the functionality desired by the customer element of the defence competence bloc and that customer’s willingness to pay for various attributes of functionality. A key element of such innovative activity in defence sectoral innovation systems is the identification of promising domains of application of existing technologies. In defence sectoral innovation systems, however, the identification of such domains is often quite uncertain, so that the creative link of technology to application domain is a quintessential entrepreneurial activity.132 The Swedish and Australian case studies show how innovation performance is affected by the willingness of defence actors to accept, and by their ability to encourage, such entrepreneurial activity.
Access by innovators and entrepreneurs to established technologies is a prerequisite for technological speciation along the above lines. Such access is typically a consequence of the diffusion of established technologies. In defence sectoral innovation systems, however, the diffusion of military technology, and of civil technology with military applications, is controlled by national governments and is not solely determined by market forces.
National defence actors in technology leading nations that have invested heavily in development of innovative solutions to military capability requirements in a quest for military advantage will have a strong incentive to control the diffusion of that technology. In these circumstances the nature and scale of military technological diffusion will be mediated by a mix of market and policy institutions. Specifically, technology leaders will have a compelling incentive to deny adversaries access to any technology with the potential to undermine their military technological advantage. Conversely, technology leaders will have more incentive to share military technology with friends and allies with a view to encouraging those friends and allies to share the burden inherent in competing for military advantage over a common adversary.
131 David Levinthal, The slow pace of rapid technological change: gradualism and punctuation in technological change, Industrial and Corporate Change, Vol 7 (2), 1998, pp. 217-220. 132 ibid., page 220. 50
The incentive for technology followers to access leaders’ technology (that is, to exercise ‘buy’ choices) will depend on the acceptability to defence actors of any extra cost, schedule and technical risk inherent in ‘make’ choices over ‘buy’ choices. From a technology follower’s perspective, the relative merits of such ‘make’ or ‘buy’ choices (and, by extension, the technological content of innovation outcomes) is also affected by the terms and conditions imposed by technology leaders as part of the ‘deal’ between leader/supplier and follower/customer. In general, however, by permitting the diffusion of their innovations, technology leaders can reduce the incentive for technology followers to exercise ‘make’ choices. To this extent, managed diffusion can reduce the tendency for followers to proliferate alternative solutions to military capability requirements and constrain the plurality of technological choice.133 The Swedish and Australian case studies show how innovation performance is affected by the judgements actors make regarding the relative merits of ‘make’ or ‘buy’ choices in providing novel solutions to requirements for military capability.
3.2.5 Demand Demand constitutes the fifth and final building block of defence sectoral innovation systems. A decision-maker triggers demand by deciding to act on a requirement for military capability. In this thesis, demand relates to action taken by the customer element of the defence competence bloc, firstly, to search for a solution to a requirement for military capability; secondly, to select a solution from the menu of candidate solutions so identified; and thirdly to procure the solution so identified. These three components of the demand building block are described in more detail in the following paragraphs.
In defence sectoral innovation systems, the search process is initiated when actors performing the customer function deem the requirement for one military capability sufficiently compelling to command priority over other requirements in the competition for finite defence resources. The search process is conducted by actors performing the customer function in the defence competence bloc. Because these actors are boundedly rational, they cannot choose among all possible options and can only scan a few alternatives in a limited search space. This thesis follows McKelvey in defining ‘search space’ in terms of economic and technological opportunities. Accordingly, in defence sectoral innovation systems economic opportunities are created when a requirement for military capability attracts budget support by the national government. Technological opportunities are inherent in the portfolio of technologies that actors can use or adapt to solve the problems identified in meeting the demand for military capability.134 The Swedish and Australian case studies show how innovation performance was affected by the efficiency and effectiveness
133 Peter Hall and Robert Wylie, Arms export controls and the proliferation of military technology in B. Goldsmith and J. Brauer (eds), Economics of War and Peace: Economic, Legal, and Political Aspects, Emerald Group Publishing, Bingley, UK, 2010, especially pp. 66-67. 134 McKelvey, pp. 209-216. 51
with which actors performing the customer, innovator and entrepreneur functions were able to identify economic and technological opportunities identified in the search for novel solutions to capability requirements.
In defence sectoral innovation systems, the selection process is also conducted by actors performing the customer function in the defence competence bloc. In selecting an innovative solution to a requirement for military capability within the constraints of limited resources, the actors involved will trade off various functional attributes and the cost and risk involved. This thesis analyses the trade-offs involved in terms of value for money. The concept of value for money provides a useful way of assembling the various components of defence demand into a coherent frame of reference. As indicated in Figure 3.1, value for money in current British usage is about striking a balance appropriate to the particular case between economy (that is, spending less on inputs), efficiency (that is, a measure of productivity or output relative to input) and effectiveness (that is, a measure of the impact achieved).135
Figure 3.1 ‘Value for Money’ in British usage136
A distinctive feature of judgements about the relative value for money of competing solutions to military capability requirements is the paucity of information conveyed by price. In these circumstances, judgements about value for money in defence innovation are informed by assessments of military capability value rather than market-mediated prices. A distinctive feature of military technological innovation is that its true effectiveness may only be demonstrated in combat and only be fully apparent after the event.137 It follows that different actors in different defence sectoral innovation systems will have widely varying views about the appropriate balance to be struck between economy, efficiency and effectiveness in any particular case. The Swedish and Australian case studies show how
135 See http://www.idea.gov.uk/imp/core/page.do?pageId=1068398 accessed 14 November 2008. 136 ibid. 137 See Stefan Markowski, Peter Hall and Robert Wylie, Procurement and the chain of supply in Stefan Markowski, Peter Hall and Robert Wylie (eds), Defence Procurement: A Small Country Perspective, Routledge, 2010, pp. 14-16. 52
innovation performance is affected by the efficacy of the institutional and other arrangements for reconciling such varying views.
In defence sectoral innovation systems, the procurement process is the process by which the actors performing the customer function in the defence competence bloc acquire the goods and services identified via the above search and selection activity. In democracies, however, defence procurement processes are embedded in the nation-specific institutional arrangements discussed earlier. Such processes will therefore constitute distinctive features of defence sectoral innovation systems.
In addition, procurement processes will reflect nation-specific institutional arrangements and, to that extent, will help distinguish one defence sectoral innovation system from another.138 The Swedish and Australian case studies show how differences in procurement processes can lead to differences in innovation performance. A threshold difference relates to the efficiency and effectiveness with which the actors performing the customer and the industrialist functions in the defence competence bloc exchange information about what military capability the customer requires. A second difference relates to how the customer defines, and secures, value for money (including the role of competition in optimising that value). A third difference relates to how those actors identify and share the risk inherent in developing novel solutions to capability requirements. A fourth difference relates to the contracting arrangements between customer and industrialist.
Section 3.3 Conclusion The above framework is intended to facilitate investigation of why Sweden and Australia – two nations at comparable stages of economic development, with comparable democratic political systems and with access to comparable technologies – performed differently in generating novel solutions to their respective requirements for broad area surveillance capability. The framework highlights several subsidiary research questions which will be pursued in the case studies of the innovation performance of the Swedish defence sectoral innovation system (Chapters 4-6) and of the Australian defence sectoral innovation system (Chapters 7-9). These subsidiary research questions, which flow from the five defence sectoral innovation system building blocks, are summarised in the following paragraphs.
The first subsidiary research question concerns Swedish and Australian institutions. These include, firstly, national security policy and/or grand strategy; secondly, the narrative used by governments to justify to electors and taxpayers the allocation of finite resources to military technological innovation; and, thirdly, the governance arrangements that guide the management of the resources allocated to military technological innovation. The subsidiary question can be stated as: “What is distinctive about the Swedish and Australian institutions and how did those distinctive institutions affect each country’s development of novel
138 ibid., pp. 68-71. 53
solutions to their respective requirements for a broad area surveillance capability?” Accordingly, Chapters 4-6 will consider how particular Swedish institutions affected the time Swedish actors took to produce a novel solution to the SwAF requirement for a broad area surveillance capability, the cost they incurred in doing so and the pattern of diffusion of that solution after its acceptance into service. Similarly, Chapters 7-9 will consider how particular Australian institutions affected the time Australian actors took to produce a novel solution to the Australian defence requirement for a broad area surveillance capability, the cost they incurred in doing so and that solution’s pattern of development after it was accepted into service.
The second subsidiary research question concerns Swedish and Australian actors and networks. As explained earlier in this chapter, this thesis uses Eliasson’s notion of a competence bloc to frame a discussion of the roles played by various actors, and the impact of the networks among those actors, in defence sectoral innovation systems. On this basis, the second subsidiary research question can be stated as: “How did Swedish and Australian actors perform the defence competence bloc functions of informed customer, innovator, entrepreneur, venture capitalist and industrialist and how did the performance of these functions affect each country’s development of novel solutions to their respective requirements for a broad area surveillance capability?” Accordingly, Chapters 4-6 will consider how performance by Swedish actors of the defence competence bloc functions affected the time they took to produce a novel solution to the SwAF requirement for a broad area surveillance capability, the cost they incurred in doing so and the pattern of diffusion of that solution after its acceptance into service. Similarly, Chapters 7-9 will consider how performance by Australian actors of the defence competence bloc functions affected the time they took to produce a novel solution to the Australian defence requirement for a broad area surveillance capability, the cost they incurred in doing so and that solution’s pattern of development after it was accepted into service.
The third subsidiary question concerns Swedish and Australian military doctrine which is about how military actors use military artefacts to apply military force to achieve a military effect. The third subsidiary question can be stated as: “What are the distinctive features of Swedish and Australian military doctrine and how did those features affect each country’s development of novel solutions to their respective requirements for a broad area surveillance capability?” Accordingly, Chapters 4-6 will consider how Sweden’s ground- based air defence doctrine affected the time actors in the Swedish defence competence bloc took to develop a novel solution to the SwAF requirement for a broad area surveillance capability, the cost they incurred in doing so and the pattern of diffusion of that solution after its acceptance into service. Similarly, Chapters 7-9 will consider how Australia’s core force and other doctrines affected the time taken by actors in the Australian defence competence bloc to develop a novel solution to the Australian defence requirement for a
54
broad area surveillance capability, the cost they incurred in doing so and that solution’s pattern of development after it was accepted into service.
The fourth subsidiary question concerns the Swedish and Australian technology bases. Each country’s technology base reflected the technology paradigm prevailing among the actors populating the respective defence sectoral innovation systems. Those technology paradigms fostered the development of distinctively configured technological systems which developed along distinctive trajectories. Hence the fourth subsidiary question can be stated as: “What are the distinctive features of the Swedish and Australian technology bases and how did those features affect each country’s development of novel solutions to their respective requirements for a broad area surveillance capability?” Accordingly, the case studies in Chapters 4-6 will analyse how the distinctive features of Sweden’s technology base affected the time actors in the Swedish defence competence bloc took to develop a novel solution to the SwAF requirement for a broad area surveillance capability, the cost they incurred in doing so and the pattern of diffusion of that solution after its acceptance into service. Similarly, the case studies in Chapters 7-9 will analyse how the distinctive features of Australia’s technology base affected the time actors in the Australian defence competence bloc took to develop a novel solution to the Australian requirement for a broad area surveillance capability, the cost they incurred in doing so and the pattern of diffusion of that solution after its acceptance into service.
The fifth subsidiary question concerns Swedish and Australian demand, defined in terms of the search for, selection of and procurement of novel solutions to their respective capability requirements. Hence the fifth subsidiary question can be stated as: “What are the distinctive features of the processes by which Swedish and Australian actors exercised demand for solutions to capability requirements and how did those features affect each country’s development of novel solutions to their respective requirements for a broad area surveillance capability?” Accordingly, the case studies in Chapters 4-6 will analyse how the distinctive features of Sweden’s demand processes affected the time actors in the Swedish defence competence bloc took to develop a novel solution to the SwAF requirement for a broad area surveillance capability, the cost they incurred in doing so and the pattern of diffusion of that solution after its acceptance into service. Similarly, the case studies in Chapters 7-9 will analyse how the distinctive features of Australia’s demand processes affected the time actors in the Australian defence competence bloc took to develop a novel solution to the Australian requirement for a broad area surveillance capability, the cost they incurred in doing so and the pattern of diffusion of that solution after its acceptance into service.
55
Chapter 4: Sweden’s defence sectoral innovation system This thesis uses case studies of military technological innovation in Sweden and Australia to determine why similar countries tend to diverge in terms of the time they take, the cost they incur and the patterns of diffusion they display in addressing similar military requirements. This is the first of the three chapters of the thesis that describe, respectively, the structure, conduct and performance of the Swedish defence sectoral innovation system. It describes the structure of the Swedish defence sectoral innovation system in terms of the generic framework for such systems set out in Chapter 3. Accordingly, the chapter is organised around the following defence sectoral innovation system building blocks: Section 4.1 describes the institutions that conditioned innovation choices by Swedish actors in distinctive ways. Section 4.2 identifies the actors that populated the Swedish defence sectoral innovation system and describes how they performed the defence competence bloc functions in generating novel solutions to Swedish requirements for military capability. Section 4.3 describes the distinctive doctrine of ground-based air defence that shaped military technological choices by those actors. Section 4.4 describes that element of the Swedish technology base that Swedish actors drew on in developing a solution to Sweden’s requirement for a rapid reaction surveillance capability. Section 4.5 describes how those actors searched for, selected and procured a novel solution to Sweden’s requirement for a rapid reaction surveillance capability. Section 4.6 concludes the chapter.
Section 4.1 Swedish Institutions During the Cold War, norms that shaped the structure and operation of Sweden’s defence sectoral innovation system included, firstly, Swedish armed neutrality (which dates from the 1830s and the nation’s response to defeats by a resurgent Russia). The second norm was the socio-political rationale for Swedish corporatism (which dates from the late 1930s and the nation’s response to World War Two). The third distinctive norm was Swedish governance (which entails the separation between strategy formulation and strategy implementation by technical and administrative experts, an arrangement dating from the 1630s).
Swedish armed neutrality comprises political, military and technological elements. The political element of Swedish national security policy entailed convincing its neighbours, primarily Russia and later the Union of Socialist Soviet Republics (USSR), that it would remain neutral in any war and, thereby, reduce those neighbours’ incentive to attack it. To this end, Sweden consciously avoided entering into military alliances in peacetime. Despite declaring its neutrality in World War One, Sweden was adversely affected by trade embargoes. In response, the Swedes formed a government of national unity which, by 56
encompassing a broad spectrum of Swedish political and economic interests, prepared the way for the distinctive corporatist arrangements that Sweden established during and after World War Two.
During the 1920s, commitment to disarmament and the establishment of the League of Nations reduced Swedish governments‘ incentives to invest in military capacity.139 Swedish companies like Bofors offset reduced domestic demand for munitions by exporting. This legitimised the commercial supply of military equipment and established a pattern of commercial leadership of Swedish military technological innovation that prevailed during and after the Second World War. Sweden began rearming in 1936 but the prevailing pre- war tensions effectively prevented it from importing combat aircraft.140 When World War Two broke out Sweden declared its neutrality but continued trying, ultimately unsuccessfully, to import combat aircraft. During World War Two Swedish defence planners sought to convince any would-be invader that any attempt to occupy Sweden would be prohibitively expensive, particularly if that invader was already fighting elsewhere. To this end the Swedes sought to complement land and maritime forces with a credible air defence capability based on limited imports of German aircraft (see below), construction of overseas aircraft designs by local industry and, when the latter encountered difficulty, local construction of indigenous designs.
Nazi Germany’s recognition of Swedish neutrality was subject to Swedish compliance with German wishes. For example, in April-June 1940 Nazi leaders stopped delivery of materiel (notably German combat aircraft) in order to pressure Sweden to allow German troops and supplies to transit Swedish territory during the fight for the Norwegian port of Narvik. Sweden’s experience of materiel-based policy coercion and sovereignty constraint in the lead up to, and during, World War Two, helps explain the high value Swedes accorded to indigenous ‘make’ solutions to materiel requirements during the Cold War.141
During World War Two Sweden again established a government of national unity. The Swedish Air Force worked with Swedish companies to establish the Swedish aircraft industry needed to supply combat aircraft unavailable from overseas. This laid the foundations for Sweden’s Cold War aircraft industry and the associated air defence technology base. In order to facilitate the mobilisation of Swedish industry during the Second World War, Sweden’s government of national unity also convened the National Industry Committee (IK). The IK was chaired by prominent industrialists and, in managing Swedish industrial mobilisation, exercised authority delegated to it by the Swedish government. The Swedes rescinded their government of national unity after the end of the Second World War but
139 Nils Gylden, interview, Stockholm, 15 July 2010. 140 Ulf Olsson, The Creation of a Modern Arms Industry: Sweden 1939-1974, Goteborg, 1977, pp. 29-56. 141 ibid., p. 48. 57
post-war geopolitical circumstances prompted them to retain IK-inspired corporatist relationships between government and industry for the duration of the Cold War.142
By the end of the Second World War the Soviet Union had established undisputed control over the Baltic coast from Finland almost to Denmark. Swedes considered it their only credible threat and, neutrality notwithstanding, saw their interest best served if the West prevailed in the event of a war between the East and Western blocs. This preference biased Sweden in favour of advanced Western military technology during the Cold War.143 In favouring advanced Western technology, however, Swedish defence actors risked compromising Sweden’s political need to appear credibly neutral during the Cold War. The dilemma became increasingly acute as the Cold War progressed. As the 2002 Security Policy Enquiry stated:
Even though Sweden could assemble a large army, it would obviously have been difficult to offset, in quantitative terms, almost a dozen divisions located in the Leningrad and Baltic military districts. For Sweden it therefore became natural to try to equip and organise its forces with technical superiority in mind ... it was evident that if the Swedish Armed Forces were to be able to uphold a superior technical level, Sweden would have to purchase advanced military technology abroad. In the post- war period, the United States, by far the leading military industrial power, had built up considerable military technological capacity. It became a vital interest for Sweden to develop close contacts and cooperation with the United States, particularly in regard to weapon research and development and other military technology.144
Particularly in the Cold War period, however, US and other Western governments exercised stringent policy control over the diffusion of the military technology Sweden sought. The US government exercised this control by requiring US companies selling to Sweden to obtain export licences issued by government officials on a case-by-case basis. Swedish compliance with US government export licensing conditions risked compromising the Swedish government’s case that Sweden was neutral, thereby enhancing the government’s preference for indigenous ‘make’ solutions to Swedish military requirements. It also encouraged Swedish companies to invest heavily in upstream R&D capacity. Swedish air defence doctrine favoured dispersed air force units operating advanced platforms and systems from hardened bases supported by conscripted personnel with limited technical training. This required a kind of ‘advanced simplicity’ that was not available from the
142 ibid., p. 59. 143 Lennart Kallquist, interview, Stockholm, 14 November 2009, p. 7. 144 Inquiry on Security Policy, Peace and Security: Swedish Security Policy 1969-1989, Abridged Version and Translation of SOU 2002:108, Statens Offentliga Utredningar, Stratsredsberedningen, Stockholm, 2004, p. 101. 58
countries from whom Sweden was prepared to buy.145 This aspect of Swedish military doctrine also favoured ‘make’ solutions to military capability requirements.
During the Cold War, Swedish domestic political imperatives also favoured such ‘make’ solutions: workers in the Swedish military aircraft industry were represented by the strongly anti-communist Swedish Metal Workers Trade Union which sought to counter communist influence by maximising the opportunities for high-value work for its members. Hence the union and union-affiliated members of the Riksdag supported the procurement of locally designed and developed military aircraft.146
The second Swedish norm shaping Swedish military technological innovation during the Cold War was the pervasive Swedish socio-political consensus in favour of corporatist arrangements for the supply of materiel, including, in particular, that required for ground- based air defence. Corporatism is defined here as the institutionalised participation by private interests in the formulation and implementation of public policy. Corporatism permits private interests to share in the sovereignty of the state and therefore involves more than mere consultation by the state with interest groups.147 Swedish corporatist arrangements emerged in the late 1930s, became a distinctive feature of the Swedish defence innovation system during the Second World War and continued to influence the structure, conduct and performance of Swedish defence innovation during the Cold War. For example, in the mid-1930s, rival Swedish entrepreneurs responded to the Swedish government’s efforts to procure combat aircraft by establishing competing enterprises to build German and US aircraft designs under licence.148 In December 1936, with the support of the Swedish government of the day, the Commander in Chief of the Swedish Air Force suggested the two aircraft groups merge to make best use of the nation’s scarce engineering resources. This government initiative prompted protracted commercial negotiations culminating, in February 1939, in the formation of the Svenska Aeroplanaktiebolaget (SAAB), whose Board was soon dominated by the remarkable Marcus Wallenberg (see below).
In 1940 SAAB (and other Swedish aircraft component manufacturers) negotiated the Framework Agreement with the Swedish Air Ministry. The Framework Agreement (which remained in effect until the 1960s) provided for a cost-plus contracting arrangement which created the commercial incentive for Swedish companies to undertake the design, development and production of combat aircraft and other materiel on a corporatist basis. Prices were determined when manufacturing was complete and the costs of materials and
145 Kallquist, interview, p. 7. 146 Gylden, interview, p. 3. 147 F. Van Waarden, Crisis, corporatism and continuity in W. Grant, J. Nekkers and F. Van Waarden (eds), Organising Business for War: Corporatist Economic Organisation During the Second World War, Berg, Oxford 1991, pp. 12-14. 148 This section draws heavily on Ulf Olsson, Furthering a Fortune, Marcus Wallenberg Swedish Banker and Industrialist 1899-1982, Ekerlids Forlag, Stockholm, 2001, pp. 292-298. 59
wages known. The Agreement defined provisions for depreciation and write-off of capital, plant and equipment and stipulated a profit of 10% of the work’s cost price.149
During the Second World War, Wallenberg and his associates helped maintain the political legitimacy of the Framework Agreement and of the associated corporatist arrangements by fostering close links with government agencies, including the IK, responsible for procuring military equipment. Close wartime links were based on transparent transactions and were designed to facilitate the allocation of scarce resources among the companies involved in Sweden’s war effort. Those wartime links prepared the way for the dense networks among the actors in the Swedish defence competence bloc that characterised the Swedish defence innovation system during the Cold War. Even after the Agreement fell into desuetude in the 1960s, the Swedish consensus in favour of corporatist arrangements endured for as long as the Soviet threat lasted. In the 1980s, for example, corporatist arrangements underpinned the structure and operation of the Industrigruppen JAS (the JAS Industrial Group) created by Marcus Wallenberg to develop a replacement for the obsolescent Viggen fighter aircraft.150
The third Swedish norm shaping Swedish military technological innovation during the Cold War was the governance arrangements underpinning the conduct of Swedish defence business. Swedish innovation outcomes were particularly influenced by two aspects of Swedish governance. The first aspect was the autonomy with which executive agencies like FMV (responsible for procurement – see below) performed their functions. The second aspect related to the relatively loose arrangements by which the Riksdag held the Ministry for Defence accountable for the use of resources appropriated for Defence in the annual budget process.
The arm’s length relationship between elected Ministers (who are accountable to the Riksdag for the formulation of policy) and the expert non-elected agencies (who implement the policies for which the Minister was accountable to the Riksdag) encouraged the agencies to accept the risk inherent in devising novel solutions to capability requirements. These arrangements derived from the response to King Gustav II Adolf’s ill-informed intervention in the design and construction of the Wasa warship, leading to the loss of that vessel in 1628.151 In 1630 Axel Oxenstierna, the King’s Chancellor, separated policy and strategy (the Swedish monarch’s prerogative) from production and procurement of the materiel needed to execute the monarch’s strategy and established the Royal War Materiel Board to foster the requisite technical and administrative expertise. This institutional separation between policy formulation and policy implementation was formalised in the 1809 Swedish
149 Olsson, The Creation of a Modern Arms Industry, p. 123. 150 Olufsson, Furthering a Fortune, p. 297. 151 Jan Glete, Swedish Naval Administration 1521-1721: Resource Flows and Organisational Capabilities, Brill, Leiden, 2010, pp. 287-292. 60
Constitution.152 It underpinned subsequent arrangements for Swedish military procurement which culminated, in 1968, in the establishment of the Forsvarets materielverk, or Defence Materiel Administration (FMV).
FMV and similar Swedish government agencies act independently in realising the goals of government within the framework of its policy guidance and subject to the resources appropriated by the Riksdag and allocated by the executive. The Swedish Minister for Defence and officials of the Swedish Ministry of Defence are not allowed to intervene in the day-to-day operation of FMV or in the handling of specific procurements. FMV is, however, obliged to obtain ministerial endorsement of the guidelines they envisage following when implementing a government decision.153 Within these limits, however, the autonomy of FMV and its predecessors fostered an informed appetite by the officials involved for innovation risk. The Swedish customer’s informed risk appetite accommodated business mistakes by Swedish innovators and entrepreneurs, encouraged them to learn from those mistakes and enabled them to husband the knowledge they gained from such learning.
During the Cold War, the influence of FMV autonomy on Swedish military technological innovation was enhanced by the loose fiscal accountability that characterised Swedish budget procedures at that time. In principle, democratic resource allocations will be biased towards spending when the benefits of spending can be targeted at particular constituencies while the costs are distributed over a broad spectrum of taxpayers.154 Accordingly, members of the Riksdag Defence Committee were inclined to favour defence spending, not only because of the perceived Soviet threat but also because under the prevailing Riksdag rules the Committee was not confronted with the full cost of its decisions. During the Cold War, overall, the prevailing financial management arrangements imposed limited financial constraints on the Swedish defence customer working with Swedish innovators and entrepreneurs on high-risk/high-return solutions to Swedish military requirements.
These loose fiscal accountability arrangements prevailed during the Cold War, up until the mid-1990s. Indeed, Riksdag institutional arrangements enabled the Riksdag itself to avoid confronting the costs of its decisions in appropriating funds for both defence and non- defence national priorities. Loose fiscal accountability arrangements favoured major Swedish air defence innovations like the expensive and controversial Viggen and Gripen development programs but helped cause Sweden’s fiscal crisis in the early 1990s, just as the
152 Jon Pierre, Legitimacy, institutional change and the politics of public administration in Sweden, International Political Science Review, Vol 14 (4), 1993, p. 390. 153 Swedish Institute Fact Sheet: The Swedish System of Government, p. 4, available at http://www.sweden.se/upload/Sweden_se/english/factsheets/SI/SI_FS55z_The_Swedish_System_of_Governme nt/The_Swedish_System_of_Government_FS55z.pdf, accessed 1 February 2010. 154 Joachim Wehner, Budget reform and legislative control in Sweden, Journal of European Politics, Vol 14 (2), p. 315. 61
Cold War was ending.155 The fiscal crisis prompted far-reaching reforms of Riksdag budgetary institutions. These reforms entailed more stringent scrutiny of Swedish defence proposals for high-risk/high-return development projects than was the case under pre-1996 budget arrangements.156
Section 4.2 Swedish actors, networks and the Swedish competence bloc This section analyses how Swedish defence innovation was influenced by the way Swedish defence actors performed defence competence bloc functions. These functions comprised, firstly, the informed customer, which involved judging the value of an innovation and being willing and able to pay for it. The second defence competence bloc function was the innovator, which involved addressing problems through a combination of old and new technologies. The third function was the entrepreneur, which involved identifying an opportunity by assessing, ex ante, the innovative technological combinations most likely to best meet a requirement for military capabilities and then acting on that opportunity by marshalling the resources required to realise it. The fourth function was the venture capitalist, which involved not only financing early start-ups but also judging, ex ante, which entrepreneurs were most likely to succeed in realising an opportunity to meet a requirement for military capability. The fifth defence competence bloc function was the industrialist, which entailed realising the solution to a requirement by undertaking the requisite production, marketing and distribution.
In the Swedish defence sectoral innovation system, various aspects of the informed customer function of the competence bloc were performed by interlinked Ministers, planners, enablers, procurers and military users. The Swedish Minister for Defence sets the parameters for military technological innovation by convening Commissions to review Swedish defence policy and by allocating resources to the expert agencies responsible for implementing that policy.157 As authoritative guides to Swedish defence policy, Commission Reports constitute a key institution that links members of the customer element of the Swedish defence competence bloc to each and to other elements of that bloc.158 During the Cold War, these Commissions reflected the political composition of the Riksdag, consulted widely and formulated evidence-based policy recommendations. When a Commission recommended a course of action that required the commitment of resources and the Ministry of Defence and the relevant expert agencies endorsed that course of action, then provision for the requisite resources was made in the government’s budget bills and submitted to the Riksdag for approval on a case-by-case basis. This distinctive policy process
155 Wehner, p. 319. 156For a detailed analysis see Jon Blondal, Budgeting in Sweden, OECD Working Papers, Vol VI, no 47, OECD, Paris. 1998, pp. 12-19. 157 Ministry of Defence, This is the Ministry of Defence, pp. 18-19, available at http://www.sweden.gov.se/content/1/c6/10/66/84/36f8afde.pdf, accessed 1 February 2010. 158 Kjell Goldmann, The Swedish model of security policy, West European Politics, Volume 14 (3), 1991, p. 141. 62
did not eliminate disputation among Swedish defence actors during the Cold War.159 However, the process did foster a consensus among Swedish defence actors as to the requirements for military capability, the priority for solution to those requirements and the broad parameters within which a solution would be sought. This consensus, once established, enabled Swedish actors to develop novel solutions to Swedish requirements more economically, efficiently and effectively than would otherwise have been the case.
The Swedish Minister for Defence is also accountable to the Riksdag for ensuring the expert agencies for which he or she is responsible (including the Headquarters of the Swedish Armed Forces, FMV and the Swedish Armed Forces) manage the development and deployment of Swedish military capabilities within the legislative and resources parameters approved by the Riksdag. To this end, the Minister for Defence issues letters of instruction prepared by officials in the Ministry of Defence after the Riksdag has finalised the annual Budget and returned it to the Executive for implementation. In such letters of instruction the Ministry stipulates both operational and financial objectives for the agency for the forthcoming financial year. The close accountability of Swedish Defence Ministers for defence policy outcomes, combined with the considerable autonomy accorded Swedish expert agencies in generating the outputs required to achieve those outcomes, seems to have encouraged organisational learning that enabled actors in the Swedish defence competence bloc to meet requirements more cheaply and quickly than would otherwise have been the case.
Swedish defence planners identified requirements for new military capabilities. They also specified the performance and other standards that the solution needed to meet in order to be militarily competitive. Finally, they determined the value accorded to a solution and, thereby, established the relative priority of solutions to various requirements in the competition for finite resources. During the Cold War this function was centralised under the Supreme Commander of the Swedish Armed Forces, who was supported by a Headquarters staff populated by personnel posted in from the individual armed services. Service staff in the Headquarters initiated the innovation process by undertaking the force- on-force planning to inform requirements for military capability. They influenced the innovation process by adjudicating inter-service competition for resources through the development and management of an integrated rolling program of investments to address those requirements.160 The capability investment program provided the basis on which Headquarters staff task FMV and other specialist government agencies. In turn, such tasking set the parameters for Swedish military technological innovation.
159 See I. Dorfer, System 37 Viggen – Arms, Technology and the Domestication of Glory, Universitetsforlaget, Oslo, 1973, pp. 175-184; and W.J. Taylor, Sweden in W.J. Taylor and P. Cole: Nordic Defense: Comparative Decision Making, Lexington Books, Lexington, 1985, pp. 166-179. 160 Cedric Pugh, Sweden’s approach to improve effectiveness in Public Administration 1967-86, Financial Accountability and Management, Vol 4 (1), Spring 1988, p. 53. 63
Swedish defence enablers provided other elements of the customer group with the specialist scientific expertise. The Swedish customer relied on in-house defence research establishments for specialised scientific input into formulating requirements, engaging innovators and entrepreneurs in developing novel solutions and in judging the relative merits of candidate solutions. During the Cold War, such in-house scientific advice was provided by, for example, the Swedish Institute for Aeronautic Research (Flygtekniska försöksanstalten – FFA), established in 1940 and the Swedish Defence Research Establishment (Försvarets forskningsanstalt – FOA), established 1945.161 In 2001 these organisations were combined into a single Defence Research Institute (FOI) and their programs aligned more closely with post Cold War priorities.
During and after the Cold War, FOI and its predecessors collaborated closely with Swedish universities and polytechnics as well as Swedish industry (which had developed very substantial in-house R&D capacity – see discussion of industrialists below). Such collaboration, which was expressly encouraged by Swedish legislation,162 was designed to fill residual needs not covered by either Swedish universities or by Swedish defence companies. During the Cold War, this approach was made possible by the dense networks linking the Swedish customer, innovators in Swedish universities and innovators in Swedish companies to each other.
Swedish defence procurers searched for solutions to the military capability requirement identified by the planners, selected solutions from the candidates identified in the search process and procured the solution so selected. Within the customer element of the Swedish defence competence bloc, FMV is responsible for managing the procurement function. In doing so it establishes the link between requirements for military capability (formulated by the Swedish planners) and the generation of novel solutions to that requirement (by the innovators, entrepreneurs and industrialists that populate the rest of the Swedish defence competence bloc). FMV established this link, which drove the Swedish defence innovation system, by procuring artefacts from industrialists. The latter define Swedish innovation outcomes by designing, developing and producing artefacts that realise the opportunities identified by entrepreneurs and incorporate the novel combinations of technology devised by the innovators.
FMV influences the Swedish defence innovation process by judging the relative value for money of candidate solutions to capability requirements. Such judgements include, firstly, assessing the degree to which the artefact designs proposed by industrialists fit the requirements; secondly, assessing the degree of cost, schedule and technical risk inherent in
161 See FOI Annual Report 2001, Stockholm, p. 3, available at http://www.foi.se/upload/omfoi/infomaterial/foi- annual-report-2001.pdf, accessed 3 February 2010. 162See Goran Marklund, STI OUTLOOK 2002 – Country Response to Policy Questionnaire (Sweden), p. 9, available at http://en.wikipedia.org/wiki/Swedish_Defence_Research_Agency, accessed 3 Feb 2010. See also Bo Tarras-Wahlberg (Director of Research Strategy and Markets, FOI), interview, Stockholm, 17 November 2009. 64
the artefact designs proposed by industrialists; and, thirdly, gauging the appropriate allocation of that risk between customer and industrialist. In making these judgements during the Cold War, FMV drew on a combination of technical expertise, deep knowledge of industrialists’ capacity and extensive experience in constructing deals with those industrialists that create value for money acceptable to Ministers and the Riksdag. An indication of FMV’s deep and extensive technical competence was the qualifications of its 2000 employees, of whom some 20% have Masters of Engineering, a further 20% were graduate Engineers, 3% had Masters of Business Administration while a further 15% had other professional qualifications such as law degrees. FMV’s deep technical competence enabled it to exchange technical information about requirements and solutions with innovators, entrepreneurs and industrialists efficiently and effectively.
During the Cold War, FMV actors had ample opportunity and incentive to develop deep experience in gauging the cost, schedule and technical risk inherent in artefact designs proposed by Swedish industrialists. FMV’s opportunities and incentives for accumulating such experience derived from Swedish emphasis on indigenous ‘make’ solutions to requirements and the Swedish practice of periodically upgrading artefacts in response to evolving Soviet capabilities. These features of Swedish capability development also fostered dense networks among FMV and the innovator, entrepreneur and industrialist elements of the Swedish defence competence bloc. Those networks enabled FMV and other elements of that competence bloc to exchange risk-related information efficiently and effectively.
During the Cold War, Swedish corporatism constituted a barrier for new entrants to Swedish defence business and encouraged incumbents to specialise in particular technologies. A stable population of innovators, entrepreneurs and industrialists enabled FMV actors to develop a deep understanding of, firstly, the specialist skills and competencies of other elements of the defence competence bloc. Such stability also enabled FMV actors to identify and manage incentives for these elements to interact to generate solutions that represented value for money acceptable to Ministers and the Riksdag. Dense networks enabled FMV and other elements of the Swedish competence bloc to exchange information efficiently and effectively about what candidate solutions represented the best value for money in particular circumstances.
Swedish military users tended to be the first to identify capability gaps leading to the formulation of requirements for military capability and were responsible for generating an operational effect by embedding the procured solution into the appropriate socio-technical regime. This sub-section analyses how the SwAF, which manages Sweden’s largest and most diversified portfolio of radar systems, influenced the conduct of the customer element of the Swedish defence competence bloc. During the Cold War, the SwAF was organised and equipped to counter a perceived threat of Soviet airborne assault on the margins of a wider European war by performing airborne surveillance and air defence of all Sweden’s territory
65
and interdiction of enemy supply lines and invading forces.163In undertaking these tasks, the SwAF developed a distinctive ground-based air defence doctrine which underpinned the development of Sweden’s ground-based air defence system and drove air defence-related innovation – see below. The air defence system comprised interdependent artefacts which were the product of a series of interconnected technological systems. The perceived Soviet threat was the prime impetus for development of these systems along trajectories shaped by SwAF air defence doctrine – see below.
Within the customer element of the Swedish defence competence bloc, the formulation of requirements for new air defence capability was facilitated by the dense networks that linked the SwAF (as arbiter of Swedish air defence doctrine) to Swedish planners (as adjudicators of competing capability priorities), to FOI and its predecessors (as scientific advisers) and to FMV as procurer. Dense networks between the customer element of the Swedish defence competence bloc and the innovator, entrepreneur and industrialist elements of that bloc enabled all elements to share knowledge gained by the SwAF through learning by using air defence artefacts. Such networks facilitated an informed response by these elements of the competence bloc to Swedish customer demand for novel solutions to new air defence capability requirements.
In Sweden, most military technological innovators were located in the privately owned companies, with a small minority located in Swedish universities and a few in the government-owned FOI and its predecessors. During the Cold War, most Swedish innovators involved in providing novel solutions to the Swedish customer’s requirements for ground-based air defence capability were employed by technologically specialised Swedish companies, including SAAB (airframes), Ericsson (radars), Volvo Flygmotor (aircraft engines) and Celsius (command and control arrangements). This concentration of innovators in private companies seems to have resulted from historical happenstance reinforced by Swedish customer choices. For example, in the 1930s Swedish entrepreneurs worked closely with the SwAF in establishing SAAB to supply combat aircraft on a commercial basis. In 1940, however, when SAAB encountered protracted difficulty in meeting SwAF’s wartime requirements for aircraft, SwAF established the Aeronautical Studies Workshop (FFVS) to design and build a single-seater J 22 fighter as an interim solution. Although the FFVS built nearly 200 well-regarded J 22 aircraft during the war, the Swedish government had no intention of staying in the aircraft production business, closed FFVS when the war ended and thereafter relied on SAAB exclusively.
Similarly, Ericsson’s competence in telephone technology seems to have been the main reason why the Swedish defence customer turned to that company to produce radars for the Swedish military in the early stages of the Cold War. In response, Ericsson established its Microwave Division and assembled a team of electronics engineers who learnt to design,
163 Richard Bitzinger, Facing the Future: The Swedish Air Force 1990-2005, RAND, Santa Monica, 1991, p.18. 66
develop and produce radars for the SwAF. The evolution of SwAF requirements in response to developing Soviet capabilities gave Ericsson economic and technological incentives to recruit, train and retain engineers with the knowledge and skills required to devise novel combinations of old and new technology to meet those requirements. These incentives were reinforced by the barriers to entry created by the prevailing Swedish corporatist ethos and the high value accorded to indigenous ‘make’ solutions to radar-related requirements.
Ericsson was also encouraged to invest in radar-related innovation by the Swedish defence customer’s intellectual property policy. During the Cold War, this policy enabled Swedish defence suppliers to own the intellectual property (IP) they generated in the course of undertaking government-funded development programs. The impact of Swedish defence IP policy on innovation-related incentives was enhanced by the willingness of the Swedish defence customer to pay for development work on a cost-plus basis and, thereby, assume most of the risk inherent in devising novel solutions to Swedish capability requirements.
Ericsson, SAAB and similar Swedish defence suppliers concentrated their extensive investment in innovation-related skills and technical knowledge in the downstream activities related to design, development and production of artefacts with the novel characteristics required to meet SwAF requirements. When they required upstream scientific knowledge closer to the invention end of the innovation spectrum, suppliers worked closely with like- minded researchers located in Swedish universities. For example, the SAAB aeronautical engineers responsible for designing, developing and producing successive generations of Swedish fighter aircraft were located in Linkoping and worked closely with researchers in Linkoping University. Similarly, the Ericsson Microwave Division engineers responsible for designing, developing and producing successive generations of radars for those aircraft were located in Goteborg and worked closely with electronics researchers in Chalmers University in that city.
The resulting incentive for Swedish commercial innovators to form dense networks with their academic counterparts was reinforced by the Swedish government’s willingness to pay up to half the salary of researchers employed at the university but working full or part time at a company. Conversely, companies can participate in ‘visiting professor’ arrangements by which an individual with professorial competencies but engaged in non-university employment (for example, at a company) on a part-time basis can work at a university.164These arrangements were particularly significant in the development of electronic scanned array technology for the ERIEYE radar.
During the Cold War, the Swedish defence customer’s demand for novel solutions to requirements for military capability fostered the development of two categories of entrepreneur. The first comprised capability entrepreneurs who were located in the
164 Marklund, p. 10. 67
customer element of the Swedish defence competence bloc and who recognised an opportunity to meet a requirement for military capability through an innovative technological combination. The second category constituted commercial entrepreneurs in the private companies supplying the Swedish defence customer. Dense networks between the Swedish customer and other elements of the Swedish defence competence bloc fostered the emergence of commercial entrepreneurs who identified and acted on a commercial opportunity to meet a military capability requirement using innovative technological combinations.
Swedish governance fostered the emergence of capability entrepreneurs in both the planning and procurement components of the customer element of the Swedish competence bloc. For example, the Swedish requirement for a rapid reaction surveillance capability (needed to maintain the credibility of the SwAF’s ground-based air defence system) encouraged staff officers working in the capability planning element of the Headquarters of the Swedish Armed Forces to task FMV to search for solutions to that requirement. Such tasking prompted engineers in FMV to talk to their counterparts in Ericsson Microwave Division to ascertain what radar-based options warranted further investigation. The dense networks that characterised the Swedish defence competence bloc during the Cold War enabled the capability entrepreneurs to exchange uncodified information efficiently and effectively with each other and with innovators. Such information exchanges enabled the capability entrepreneurs to define technological opportunities iteratively and to refine assessments of the associated cost, schedule and technical risk progressively.
Demand for a succession of ‘make’ solutions gave Swedish capability entrepreneurs the opportunity not only to make business mistakes but also to learn from those mistakes and to apply that learning in new ‘make’ projects. In doing so they were able to develop the combination of codified and tacit knowledge and networking skill required to judge, ex ante, what combination of technologies seemed most likely to meet a requirement for new military capability, which commercial suppliers were likely to offer best value for money, what procurement arrangements would best align the interests and incentives of customer and supplier, which stakeholders in the customer element needed to be persuaded and in what terms. Finally, their deep technical knowledge, their business acumen and their location in institutional arrangements conducive to the efficient and effective exchange of large amounts of tacit and uncodified information enabled Swedish capability entrepreneurs to develop an informed appetite for cost, schedule and technical risk.
On the supply side of the Swedish innovation system, the involvement of commercial companies in designing, developing and producing solutions to Swedish military capability requirements fostered the development of commercial entrepreneurs. Swedish corporatist norms facilitated participation by Swedish companies in successive ‘make’ projects. Such
68
participation created opportunities for Swedish commercial entrepreneurs to develop, over the course of several projects, the combination of codified and tacit knowledge and networking skill required to judge, ex ante, what combination of technologies was most likely to meet the informed customer’s demand for novel solutions on a commercially viable basis. Hence, for example, engineers in Ericsson Microwave Division judged that a combination of established pulse Doppler radar technology and emergent electronically scanned array technology was a prospective candidate for a solution to the SwAF rapid reaction surveillance requirement.
The co-location of actors performing the innovator, commercial entrepreneur and industrialist functions in a single, technologically specialised company reduced the time taken and the cost incurred by commercial entrepreneurs required to identify and act on an opportunity within the company’s technological specialisation. Dense intra-company networks were complemented by dense inter-company company networks linking commercial entrepreneurs in one technologically specialised company to their counterparts in other similarly specialised companies. Dense inter-company networks among Swedish commercial entrepreneurs enabled them to act on opportunities at the system of system level more quickly and at lower cost than would otherwise have been the case.
Private banks have performed the venture capitalist function in Swedish industrial activity since at least the late nineteenth century. In particular, the Stockholms Enskilda Bank (founded in 1856 by Andre Oscar Wallenberg), and its associated holding companies Investor and Providentia, took a special interest in Swedish defence business from the 1930s until its merger with another private bank, the Skandinaviska Enskilda Banken, in 1971. Under the guidance of Marcus Wallenberg (1899-1982), these Wallenberg organisations directly influenced the structure and operation of SAAB, Ericsson and other Swedish suppliers of air defence-related artefacts up to and during the Cold War. According to Olsson:
As owners, they were actively involved and long term investors rather than passive and short term. They aimed for shareholdings large enough to put them in a powerful position, giving them control over important decisions, such as the choice of executive managers ... only minor shareholdings were held in most of the companies regarded as being part of the Wallenberg sphere. However, the holdings were large enough to enable the sphere to influence the course of general meetings, since no other grouping could mobilise stronger support. Majority shareholdings were held only in exceptional cases. Through subsidiary companies, sometimes involving several tiers, a hold could be maintained over a large number of companies with a relatively limited input of capital, resulting in increased ownership influence.165
165 Olsson, Furthering a Fortune, p. 321. 69
The Wallenberg Group’s influence over L.M. Ericsson dates from 1933 when the company, then a major supplier of telephone equipment in Sweden and abroad, was re-organised and re-financed in the wake of commercial difficulties. In 1953 Marcus Wallenberg was appointed Chairman of the Ericsson Board, from which position he dominated Ericsson management for decades thereafter.166 Marcus Wallenberg’s skill in selecting and appointing managers, his interest in high technology and his detailed, interventionist style of management created a prospective commercial environment in Ericsson for radar-related innovation.167
In performing the industrialist function in the Swedish military technology competence bloc, Swedish actors translated an idea or prototype into an artefact that met the Swedish customer’s requirements on a commercially viable basis. In the Swedish defence innovation system, the idea or prototype was typically conceived by an entrepreneur with the prompting of an informed customer, the stimulus of an innovator and the support of a venture capitalist. To produce such an artefact the Swedish industrialist marshalled design, development and production competencies, drawing on a combination of in-house competencies, extant competencies available elsewhere and newly developed competencies created in-house or externally.
From the end of World War Two to the end of the Cold War, the Swedish air defence system underwent five cycles of development focused on the design, development and production of five generations of Swedish combat aircraft – see below. Each of these aircraft and the associated systems and sub-systems were designed, developed and produced by a stable cluster of technologically specialised Swedish industrialists. This cluster included SAAB (airframes), Ericsson (radars), Volvo Flygmotor (aero-engines) and Celsius Tech (command, control and communication systems).
This pattern of Swedish air defence-related industrial activity was a consequence of action taken by the customer element of the Swedish defence competence bloc to match evolving Soviet capabilities for airborne assault. The action taken by the customer was shaped by the SwAF’s stable doctrine of ground-based air defence. Dynamically stable threat perceptions refracted through dynamically stable military doctrine led the Swedish customer to formulate path-dependent requirements for military capability. Swedish innovators, entrepreneurs and industrialists responded to these path-dependent requirements with path-dependent solutions, manifest in the successive generations of combat aircraft highlighted above.
Path-dependent requirements encouraged industrialists to design, develop and produce successive generations of artefacts embodying technologies that evolved along trajectories broadly consistent with those requirements. To the extent that Swedish industrialists were
166 ibid., p. 311. 167 ibid., pp. 315-330. 70
able to meet evolving requirements by path-dependent technological development, they had an incentive to configure and reconfigure their industrial competencies accordingly. During the Cold War, such incentives kept the configuration of Swedish industrial competencies broadly compatible with the competency configuration required to produce novel solutions to requirements within an acceptable timeframe. This is reflected in, for example, Ericsson’s ability to meet SwAF demand for progressively more capable airborne radars, including ERIEYE. During the Cold War, the Swedish customer also accorded high value to indigenous ‘make’ solutions. This combination created periodic opportunities for Swedish individuals and groups/teams responsible for performing the innovator, entrepreneur and industrialist functions to refresh their respective skills and competencies through learning-by-doing new tasks. Within this framework, Swedish corporatism provided both the opportunity and the incentive for Swedish industrialists to husband the knowledge gained through such learning-by-doing by the individuals and groups/teams they employed. In doing so, Swedish industrialists built up a flexible and adaptable stock of competencies that was proximate to the competencies they required to meet new, but path-dependent, requirements formulated by the customer. Such proximity enabled the industrialists to adjust relatively quickly and cheaply the stock of skills and competencies available ex ante to align with those required to meet new path-dependent requirements ex post. Hence, for example, Ericsson Microwave Division was able to adapt its extant competencies in airborne microwave radar to meet the SwAF requirement for a rapid reaction surveillance system relatively quickly and cheaply.
During the Cold War, the time Swedish industrialists took to design, develop and produce novel solutions to the Swedish customer’s demanding requirements, and the cost they incurred in doing so, was influenced by those industrialists’ tendency to specialise in certain technological systems. The tendency for Swedish industrialists to develop in-house particular competencies in specific technological systems was reinforced by the customer’s willingness to share with the industrialist the risk of technology development through cost- plus contracting. Once the Swedish customer verified an industrialist’s competencies through initial procurement, subsequent procurement created a self reinforcing cycle in which extant in-house competencies were reinforced, in turn enhancing barriers to entry by potential competitors. For as long Swedish industrialists could meet their informed customer’s demand for novel solutions, the cost of producing such novel solutions tended to reflect merely the marginal cost of adjusting the stock of competencies available in-house or externally.
The dense networks that characterised the Swedish defence competence bloc during the Cold War were one consequence of such industrialist specialisation. As the cost overruns and schedule slippage of the Gripen program demonstrate, however, dense networks did not prevent Swedish industrialists underestimating the time and cost of developing the competencies necessary to achieve occasionally radical extensions of the Swedish
71
technology base required to meet Swedish requirements. The Gripen program also demonstrated, however, the ability of Swedish industrialists to capture the benefits of the necessary but unpredictable costs of experimentation (what Potts called ‘good waste’) and to minimise what Potts called ‘bad waste’ – that is, the losses due to inefficiency and a failure to harvest the results of learning through experimentation, including the failure of an experiment.
Historically, the small size of the Swedish market gave Swedish industrialists like Ericsson a strong incentive to pursue commercial opportunities in non-Swedish markets. By extension, Ericsson and other the Swedish industrialists involved in designing, developing and producing novel solutions to Swedish air defence requirements had strong commercial incentives to apply the knowledge, skills and competencies so developed to satisfy comparable non-Swedish requirements. These incentives were enhanced by Swedish defence IP policy and by the Swedish defence customer’s willingness to support exports by Swedish defence industrialists. For example, other countries had requirements for broad area surveillance that had much in common with the Swedish demand for a low-cost but capable solution to its requirement for a rapid reaction surveillance capability. Ericsson Microwave Division developed the knowledge, skills and competencies to meet this requirement by designing, developing and producing ERIEYE for the SwAF. It was in the Swedish customer’s interest for Ericsson to continue developing ERIEYE after satisfying the SwAF’s immediate needs. After Ericsson entrepreneurs identified opportunities to export ERIEYE, Ericsson innovators and industrialists adapted ERIEYE technology to suit the specific requirements of non-Swedish customers, leading to the technology’s further development and rapid diffusion.
Section 4.3 Swedish Military Doctrine This section describes how Swedish air defence doctrine shaped Swedish investment in an evolving portfolio of air defence assets which drove air defence-related technological innovation in Sweden. As already indicated, Swedish defence actors perceived the Soviet Union as the only credible threat to Sweden after the Second World War. Swedish planners recognised, however, that there was no way Sweden could provide a credible defence against the full might of the Soviet forces. But Swedish defence actors considered the scenario of a unilateral assault by the Soviet Union on Sweden as part of a bilateral Soviet- Swedish dispute too unlikely for planning purposes. Rather, Swedish defence planning focused on scenarios involving Soviet attack on the margins of a wider war in Europe. The credibility of such scenarios derived from Sweden’s location on the northern flank of the European theatre. Swedish defence planners reasoned that, in the event of a European war between NATO and the Warsaw Pact, the Soviet Union would seek to maintain access to the Baltic and Bearings seas while NATO would seek to deny Soviet forces such access. Sweden was either close to or straddled the air and maritime routes likely to be used by both NATO and Warsaw pact forces seeking to control those seas on the margins of a European war. 72
Swedish policymakers sought military capability sufficient to deter an attack by the Soviet Union as part of this contest and, if that deterrence failed, adequate to buy enough time for Western powers to come to Sweden’s aid. The plausibility of the latter scenario hinged on judgements about NATO interest in, firstly, preventing the Warsaw Pact using Swedish airspace to attack NATO forces in the North Sea and, secondly, using Swedish territory and airspace to attack the Soviet western flank and to constrain Soviet naval options in the Baltic and Bearings seas.
The terrain of western Finland and north-eastern Sweden favours defence. Hence Swedish planners sought to make any Soviet land-based assault slow and costly by relatively modest investments in land-based military capabilities that took advantage of that terrain. Up to and during the Second World War Sweden relied primarily on its Navy to interdict forces attacking across the Baltic. The rapid evolution of aircraft technology during and after World War Two, however, rendered Swedish surface combatants increasingly vulnerable to attack by Soviet aircraft operating from land-based sites in the Baltic states, Poland and north-east Germany. The main focus of Swedish defence planning was scenarios involving Soviet airborne and maritime assault across the Baltic from Finland in the north to Denmark in the south. It was these scenarios that prompted the Swedes to invest in capabilities to interdict such assaults, including quiet diesel electric submarines optimised for operations in Baltic waters and the ground-based air defence capability, which is the focus of the Swedish case study.
By the beginning of the Cold War, the SwAF had displaced the Swedish Navy as the nation’s primary instrument for interdicting attacking forces. Sweden’s distinctive air defence doctrine dates from a 10-year plan for the development of Swedish defence capabilities prepared in 1954. This plan, which was eventually accepted in February 1958, entailed investing in a more capable SwAF and, to accommodate this, slowing investment in the Army and, especially, the Navy. 168 This decision established the trajectory for Swedish development of a cluster of air defence-related technologies, including radar, during the Cold War.
To maintain a credible air defence capability Swedes resorted to expensive high technology. As a 1994 Commission of Enquiry observed:
The national defence wanted the latest and most effective equipment at the lowest possible cost. Sweden had neither the financial nor the staff resources to single- handedly remain at the forefront of developments in all areas of military technology. Among other things, Swedish aircraft projects were dependent on an influx of foreign
168 Dorfer, pp. 32-34. 73
technology. And Sweden’s defence industry could not possibly produce, at competitive prices, all the equipment required.169
The plan envisaged a ‘layered’ defence capability, the outer layer of which comprised a projected or ‘shell’ defence intended to interdict invading forces before they reached Swedish territory. Such interdiction was the primary role of Swedish Navy submarines and of SwAF combat aircraft. The intention was to halt Soviet amphibious and airborne assault forces in the Baltic, as far from Sweden as possible. In addition, the SwAF was responsible for preventing enemy aircraft from violating or transiting Swedish airspace. If the enemy succeeded in breaching the projected defence ‘shell’, then Swedish military planners envisaged recourse to the second layer of perimeter defence aimed at denying enemy forces a beach-head or preliminary foothold on Swedish territory. This would be the prime responsibility of the Swedish Army, backed up by a comprehensive coastal defence system and supported by territorial forces and the home guard. Finally, should the enemy penetrate Sweden’s perimeter defences, Sweden envisaged switching to a third layer of territorial defence and guerrilla warfare aimed at harassing the invading force, depleting units by attrition and slowing its advance.170
In providing the air defence component of the outer shell of Sweden’s layered defence capability, the SwAF opted for a ground-controlled intercept doctrine of air defence. This doctrine was analogous to that pioneered by Britain’s Royal Air Force (RAF) in the late 1930s. In the Battle of Britain, the RAF’s Fighter Command used a combination of rudimentary radar systems and observers to detect and locate incoming Luftwaffe aircraft, ground-based command and control centres to process the target-related data so gathered, a communication system to transmit the data and directions to waiting squadrons of fighter aircraft and radio telephones to direct and coordinate fighter pilots for final intercept once they were airborne.
In accordance with this doctrine, the SwAF assembled a portfolio of artefacts configured to detect and classify incoming Soviet aircraft by a mix of ground-based and airborne sensors; maritime interdiction, ground attack, reconnaissance and air defence by manned aircraft optimised for the Swedish concept of operations; and ground-based command and control of defending aircraft (including ground-based processing of target data and communication of that data to defending aircraft). Swedish air defence doctrine and, by extension, the above tasks remained essentially stable for the duration of the Cold War. The artefacts used to perform the above tasks, however, evolved continuously in response to the ‘pull’ of Swedish planners’ perception of evolving Soviet capability for airborne assault and the ‘push’ of technological developments generated within the Swedish technology base.
169 Commission on Neutrality Policy, Had There Been a War … Preparations for the Reception of Military Assistance 1949-1969, Statens offentliga utredningar, Stratsredsberedningen, Stockholm, 1994, p. 120. 170 Bitzinger, pp. 7-11. 74
4.4 Swedish technology During the Cold War, the Swedish customer’s demand for neutral technology and the high value that customer accorded to indigenous ‘make’ solutions to Swedish requirements for a credible air defence capability prompted Swedish industrialists to establish a comprehensive technology base in Sweden. That technology base comprised a diverse range of technological systems, the configuration and development of which was shaped by the SwAF’s doctrine of ground-controlled intercept and the associated portfolio of air defence artefacts assembled by the SwAF. Hence the Swedish technology base encompassed platform technologies, including airframe and associated propulsion technologies, fostered by SAAB’s design and construction of successive generations of combat aircraft. It also encompassed sensor technologies, fostered primarily by Ericsson Microwave Division’s design and fabrication of successive generations of both ground-based and airborne radar systems. The Swedish technology base included command, control and communication technologies (including systems for processing and disseminating data collected by sensor systems). This element of the Swedish technology base was primarily fostered by Celsius who designed and built successive generations of STRILC and STRIC command and control systems. Sweden’s technology base also included munitions technologies (including sensor- guided munitions compatible with the evolving combat aircraft and produced by Bofors and SAAB).
The SwAF requirement, sustained for over four decades of the Cold War, for interoperable, mutually dependent air defence artefacts caused the indigenous technological systems supplying and supporting those artefacts to coalesce into a densely integrated air defence technology base. Within this technology base, each of the above technological systems evolved along a technological trajectory that was shaped by both exogenous pressures on the technology base as a whole and also by endogenous pressures created by changes in one or more of the above systems. The overall technology base was shaped by exogenous pressures derived from the perceived need, refracted through SwAF ground-controlled intercept doctrine, to match evolving Soviet capabilities for airborne assault. For example, the perceived need to counter Soviet low-flying aircraft created the requirement for a rapid reaction surveillance capability.
The Swedish technology base was also shaped by endogenous pressures which derived from changes in one technological system (for example platform technologies). Such changes were refracted through individual companies’ competencies (for example, SAAB in the case of platform technologies). But because the design, development and production of complex air defence artefacts encompassed multiple technological systems, companies with competencies in one technological system had to identify and form networks with companies possessing complementary competencies. This process of requirements-oriented technological networking led to changes in other technological systems. Hence, for example, SAAB-based innovators, entrepreneurs and industrialists involved in designing, developing 75
and producing a new fighter for the SwAF had to form networks with their counterparts in, say, Ericsson Microwave Division (in the sensor technological system) and in Volvo- Flygmotor in the aircraft propulsion technological system. Meeting demanding requirements for radar-based detection capability prompted Ericsson Microwave engineers to form networks with scientists and engineers based in Chalmers University while meeting demanding requirements for aircraft speed, power to weight ratio and endurance prompted Volvo Flygmotor engineers to form networks with their counterparts in overseas engine companies like de Havilland and Rolls Royce in the UK and Pratt and Whitney and General Electric in the US.
A distinctive feature of the Swedish air defence technology base during the Cold War was the tendency for Swedish companies to specialise in one air defence-related technological system and to invest in the development of deep competencies in that technological system. This tendency was partly a consequence of the Swedish defence customer’s demand for technologically advanced solutions to air defence capability requirements, that customer’s willingness to accept most development risk and Swedish corporatist norms. During the Cold War, for example, SAAB specialised in platform technologies, Ericsson’s Microwave Division specialised in radar technologies, Celsius Tech specialised in command, control and communication technologies and Bofors specialised in munitions technologies. For as long as Sweden’s Cold War norms and policy settings persisted, and for as long as the companies provided acceptable value for money in meeting the informed Swedish defence customer’s needs, this pattern of technological development posed almost insurmountable barriers to entry and became a self-reinforcing spiral. That said, the companies specialising in particular technological system responded to the customer’s demand with different mixes of technologies sourced locally and overseas.
A focus on the production of successive generations of Swedish combat aircraft provides a convenient way of tracing how the Swedish air defence technology base evolved through the interaction of, on one hand, exogenous pressures on the base as a whole and, on the other hand, endogenous pressures stemming from changes in individual technological systems. The following discussion is organised chronologically, beginning with Sweden’s deployment of the SAAB J21 in the transition from World War Two to the Cold War (1945- 1950), followed by Sweden’s development of SAAB J29 Tunnan in response to the geopolitical adjustments that characterised the early Cold War (1950-1956). Sweden introduced the SAAB 32 Lansen in 1953-1959 and augmented this aircraft with the SAAB 35 Draken during the period of intense NATO-Warsaw Pact competition (1955-1974). During the Detente between East and West, Sweden introduced the Viggen fighter (1970-1990) and, in the post Cold War period, fielded the SAAB/JAS 39 Gripen (1996).
By the end of World War Two, SAAB had established basic competencies in the design, development and production of piston-engine aircraft through a combination of licenced
76
production and indigenous designs. What was then Svenska Flygmotor was building pre-war US and German aircraft engines under licence. The transition of the Swedish technology base to meet the more demanding requirements of the Cold War was initiated by SwAF demand for a jet engine fighter. This prompted SAAB and Svenska Flygmotor to convert piston-engine SAAB J21 fighters to jet-propelled J21R fighters. While the J21R airframe was designed and built by SAAB, the jet engine was a British de Havilland design built by Svenska Flygmotor under licence. The J21R fighter had no radar.
Until about 1947 Swedish air defence was a matter for the Swedish Army who used human observers to detect incoming aircraft and ground-based anti-aircraft guns to engage them. From 1944 until 1956, human observers were augmented by British-supplied AMES Type 6 light warning radars with a range of some 70 km. Observers reported to a central Army operational cell by telephone. The cell included a SwAF officer with authority to scramble a squadron of SwAF aircraft. The SwAF assumed responsibility for airspace surveillance in 1948. In 1949 it began revamping the ground-controlled intercept capability, initially in response to the Berlin Blockade of 1948-49. SwAF use of airborne radar began with the conversion of Second World War vintage SAAB B18A bombers into reconnaissance aircraft by fitting them with surplus American AN/APS-4 intercept radars. The latter were manufactured by the US Western Electric Company and entered SwAF service in late summer 1948.171 At this stage Sweden did not manufacture radar.
In the early 1950s, in response to SwAF demand for a militarily competitive jet fighter, SAAB designed and built the SAAB J29 Tunnan. This aircraft was powered by a British de Havilland designed jet engine, again built under licence by Svenska Flygmotor. The J29 Tunnan had no radar. In June 1952 the Soviet Union shot down two Swedish aircraft. This and broader global tensions prompted the SwAF to institute a Quick Readiness Alert system underpinned by a combination of human observers and more capable radars for round-the-clock radar surveillance of Swedish airspace, complemented by fighters and reconnaissance aircraft.172 This became STRIL 50 and incorporated ground-based surveillance radars bought from Marconi (UK) and from Bendix (US). STRIL 50 included hardened SwAF sectoral operations centres linked by radio and telephone to a central SwAF command and control centre able to deploy anti-aircraft guns and scramble SAAB J29 interceptor aircraft.
Ongoing East-West tensions in Europe caused the SwAF to demand a strike aircraft with greater range than the J29 Tunnan. In response SAAB designed and built the SAAB 32 Lansen during 1953-1959. The SAAB 32 Lansen fighter bomber was powered by a British- designed Rolls Royce jet engine, again built under licence by Svenska Flygmotor. Early variants of the SAAB 32 Lansen aircraft were fitted with SwAF-specified but French-designed radars, with components manufactured by Ericsson Microwave Division and other Swedish
171 Urban Fredriksson, Airborne Radars in the Swedish Air Force, available at http://www.x- plane.org/home/urf/aviation/text/airborne_radars.html, accessed 26 August 2011. 172 Commission on Neutrality Policy, p. 117. 77
companies. Later variants of the aircraft were fitted with more capable radars designed, developed and built by Ericsson Microwave Division.
At this stage of its development, the Swedish air defence technology base could not meet the Swedish customer’s demand for sophisticated fighters and other artefacts at anything like acceptable cost or within an acceptable timeframe. The Swedish customer therefore began looking to the US and other Western countries for, among other items, radar and self-guided munitions. Although that kind of technology was manufactured by commercial companies in the US, the UK and other NATO countries, its diffusion was subject to close control by the national governments concerned. Beginning in the 1950s, the members of NATO (and, later, Japan) coordinated their respective national export controls through the Coordinating Committee (COCOM).
From the outset of the Cold War, successive US government administrations made Swedish compliance with NATO East-West trade policy a precondition for that country’s access to both military and dual-use technologies. Initial US concern over neutral Sweden’s will and ability to prevent advanced Western technology diffusing to the Soviet Union (and a desire to pressure neutral Sweden to join NATO) caused the US Truman administration to delay the issuing of export licences to US companies selling to Sweden. For example, it took Sweden over a year of intense diplomatic negotiation to gain export licences for the Bendix air defence radars designed to Swedish specifications and already paid for by the Swedes. This led Sweden and the US in June 1951 to conclude the Stockholm Agreement, in which the Swedes compromised their political demand for a neutral technology and accepted US conditions regarding third party access to US technology in return for access to US technology releasable to COCOM countries.173
The Stockholm Agreement cleared the way for the Truman administration (and its successors) to adopt a more forthcoming policy on Swedish access to US technology. Perhaps more important, however, was growing US acceptance that fostering Swedish military capabilities was in the strategic interest of the US and fellow NATO members.174 That said, US licensing conditions tended to reinforce the value accorded by the Swedish customer to indigenous ‘make’ solutions to Swedish military requirements. For example, the US initially declined Sweden’s request to discuss surface-to-air guided missile technology and only agreed to sell US air defence missiles to Sweden in January 1959, after the US had decided that it was in the US national interest for NATO countries to improve their air defences.175
173 Mikael Nilsson, Tools of Hegemony: Military Technology and Swedish-American relations 1945-1962, SANTERUS Academic Press, Sweden, 2007, p. 228. 174 Simon Moores, ‘Neutral on our side’: US policy towards Sweden during the Eisenhower Administration, Cold War History, Vol 2 (3), April 2002, p. 30. 175 Nilsson, pp. 256-356. 78
Actual release of the missiles to Sweden, however, remained subject to satisfactory Swedish security arrangements, including basic security legislation, security in government departments, including both physical and personnel aspects, and Swedish industrial security. In the case of the Sidewinder 1A missile, for example, export licences were denied until US inspectors deemed Swedish arrangements satisfactory. Only then did US policymakers authorise the disclosure of classified US information to Sweden, enabling sales of Sidewinder 1A missile to proceed. Sweden accepted these onerous conditions because Bofors’s efforts to develop an indigenous ‘make’ solution to the SwAF requirement for surface-to-air missiles took too long, cost too much and failed to meet SwAF specifications.
In 1948 the SwAF sought a supersonic jet fighter specialised in the air defence role and able to intercept the high-level bombers and fighters then being deployed by the Soviet Union. In September 1949 SAAB began designing and developing a fighter/interceptor aircraft which complemented the Lansen fighter-bomber’s air-to-ground capability. The upshot was the SAAB J35 DRAKEN aircraft which the SwAF accepted into service in 1960. The J35 Draken was a second-generation jet fighter, with integrated weapon and avionics systems and a distinctive double delta wing configuration. SAAB produced some 644 Draken aircraft in various configurations and upgrades to accommodate changing roles and to take advantage of improvements in radars, engines, avionics and weapons as they became available. By 1962 the Draken aircraft, including their sensors and communication systems, had been fully integrated into the evolved STRIL 60 combat guidance and air surveillance system.
The STRIL 60 system was more automated than its predecessor. It included ground-to-air data links between dispersed ground-based command centres and the J32 Lansen and the J35 Draken aircraft, deployed in dispersed, hardened BASE 90 infrastructure. The data link between STRIL 60 operators and Draken pilots was a critical innovation. It enabled the STRIL 60 operators to communicate target data gathered by ground-based PS 08 radars (bought from Britain in the mid-1950s) to pilots of the intercepting aircraft. The data link permitted jam-resistant ground-based control of the intercept of one primary target, with the intercepting aircraft vectored by discrete commands. Evolutionary development of STRIL 60 depended on imports of both surveillance and height-finding radars, including, for example, sophisticated pop-up, frequency agile PS-860 supplied by ITT Gilfillan (US).
The early J53-A model of the Draken were fitted with a Thomson-CSF Cyrano radar produced under licence by Ericsson and designated PS-02A in Swedish service. This radar was an interim solution that initiated an airborne radar development cycle: by 1961 it had been superseded by Ericsson’s indigenously developed PS-03A radar.176 In 1965 the latter was in turn superseded by successive generations of progressively more capable radars designed and developed by Ericsson.
176 Fredriksson, Airborne radars. 79
Between 1952 and 1957 the SwAF began considering a replacement for the SAAB 32 Lansen aircraft in the ground-attack role and for the SAAB 35 Draken in the interceptor/air superiority role. The replacement aircraft had to be capable of short take off and landing in accordance with the SwAF’s dispersed basing system, to have at least Mach 1 supersonic capability at low altitude and Mach 2 performance at high altitude, and to be easily supported by conscript personnel. In September 1962, SAAB responded with a design for the radically configured SAAB 37 Viggen. SAAB began building the aircraft in 1964, the first prototype flew in September 1967 and the first squadron (flying the ground-attack variant of the Viggen) was operational in 1972. SAAB delivered well over 300 SAAB 37 Viggen aircraft between 1971 and 1990. The SAAB 37 Viggen was a third-generation jet fighter and a contemporary of the US General Dynamics F-16 and the Soviet MIG 29 aircraft. SAAB’s use of digital computerisation and advanced system integration constituted significant innovations at the time. The aircraft was powered by a US Pratt and Whitney engine, again built under licence by Volvo Flygmotor (as Svenska Flygmotor was renamed after Volvo bought a controlling interest). In obtaining this engine technology, Volvo Flygmotor took advantage of the 1951 Stockholm Agreement mentioned earlier.
Like the SAAB 35 Draken, the Viggen was built in several variants, all linked to the STRIL 60 command and control system. To replace the Lansen in the ground-attack role, SAAB designed the AJ 37 Viggen with an Ericsson PS-37 radar, optimised for air-to-air and air-to- ground operations. For maritime strike and reconnaissance operations, Ericsson modified the PS-37 radar by giving it longer range and making provision for mission analysis (designated the PS-371/A radar). The AJ 37 Viggen was followed by reconnaissance versions of the aircraft. Adapting the Viggen aircraft for interceptor/air superiority role previously performed by the Draken aeroplane took longer and entailed a major upgrade of the SAAB AJ 37 Viggen’s avionics, including its radar. The latter had to be adapted to a change in Soviet airborne assault tactics which occurred in the 1960s: the earlier tactics focused on attack at high altitudes. As the performance of defending ground-based radar improved, the Soviet air force changed to attack at low altitudes, making use of the terrain to evade detection. Low-level airborne assault tactics not only forced the defender to fly higher in order to survey a larger area, but also forced the defender to use a radar able to look down and to distinguish the radar signal reflected by the enemy aircraft from that reflected by the ground, a distinction beyond the capacity of the conventional pulse radars. To meet this requirement for a look-down/shoot-down radar the SwAF began working with Ericsson and other Swedish electronics companies to develop a pulse Doppler radar system. The latter distinguishes a target from the background by detecting small shifts in the frequency of the return signal caused by the movement of the target relative to the stationary background.
In the US, the Eisenhower, Kennedy and Nixon administrations continued and enhanced the Truman administration’s policy of releasing advanced (but non-nuclear) US technology to Sweden. The US strategic interest in fostering a militarily competitive Swedish air defence
80
capability was sufficiently strong to insulate Swedish access to US air defence-related technology from bilateral political disputes occasioned by, for example, Swedish government criticism of US involvement in the Vietnamese war.177 The US strategic interest was even strong enough to override such bilateral security issues as the compromise of Sidewinder 1A missile operational information by Wennestrom, a SwAF officer recruited by the KGB to spy for the Soviet Union. Similarly, Swedish military and commercial interest in access to US technology was sufficiently strong to curb overzealous insistence on technologically ‘neutral’ solutions to Swedish military capability requirements.
To enable the SwAF to exploit the JA37 Viggen’s interception/air superiority capabilities, however, the STRIL 60 system needed to be upgraded. This entailed, for example, introducing large capacity digitised data processing. The resulting STRIL 90 system relied on a mix of imported systems (for example those supplied by Singer Librascope (US) and GEC Marconi (UK) and locally developed systems supplied by, for example Standard Radio Sweden. A key innovation was an enhanced ground-to-air data link that permitted the Viggen to engage one primary target and four secondary targets. Other enhancements enabled the intercepting aircraft to gain tactical advantage by relying solely on target data transmitted by the ground control or another fighter, thereby obviating the interceptor’s need to illuminate its target and, by doing so, alerting the target to its presence.
In 1980, FMV briefed the Swedish defence industry about the SwAF requirement for a small and relatively inexpensive aircraft to replace the obsolescent Viggen. In addition to being militarily competitive with the new Soviet aircraft then going into service, the replacement had to be compatible with Swedish infrastructure and able to be serviced in the field by non-professional conscript personnel.178 They envisaged the new aircraft weighing half the Viggen while carrying the same weapons load but with much improved performance. The new aircraft was to have greater operational availability than the Viggen but its in-service support was to cost only 60-65% of the Viggen. To achieve this, FMV and the SwAF envisaged replacing the Viggen with a single multi-role combat aircraft that combined the functions of intercept, ground attack and reconnaissance (Jakt, Attack, Spanning – JAS).
The FMV, SwAF and Sweden’s core aerospace companies (SAAB, Ericsson and Volvo Flygmotor) all recognised that meeting these requirements would entail development of a broad range of new technologies and competencies, with concomitant cost, schedule and technical risk. For example, to achieve the agility required to prevail in contemporary air combat, SAAB proposed a dynamically unstable aircraft design employing emergent ‘fly by wire’ technology. Partly in response to FMV’s threat to buy offshore, the companies also agreed to design, develop and build the aircraft on a fixed price basis. To this end, in 1981 they formed the Industry Group JAS. That consortium comprised Ericsson (responsible for
177 Inquiry on Security Policy, p. 99. 178 Eliasson, Advanced Public Procurement, p. 250. 81
the radar, computer systems and electronics); Saab (responsible for platform development, systems integration, aircraft control systems and delivery of a complete and functioning aircraft as specified and scheduled); Volvo Flygmotor (responsible for aircraft propulsion system); and FFV AeroTech (responsible for testing and support equipment). FMV and the Industry Group JAS concluded a contract in 1982 for the development of the platform, production of five prototype JAS 37 Gripen aircraft for testing and an initial production of 30 aircraft.179 As the program encountered technical problems, cost increases and schedule delays, it became increasingly controversial. The prototype first flew in December 1988 but crashed in February 1989 due to a software fault in the flight control system. A modified prototype first flew in May 1990, followed by flight of a prototype with an operational radar in March 1991. The first production Gripen flew in March 1993 but crashed in August that year, again due to a defective flight control system. Despite the collapse of the USSR in 1991, the contract between FMV and the Industry Group JAS was renegotiated after the crash to provide for, among other things, a revised delivery schedule. In September 1993 the Minister for Defence agreed to increase the JAS 39 Gripen budget by over 20%.180 The Gripen finally achieved initial operational status in 1995 and full operational status with the SwAF in 1997.181
To meet the demanding performance specified for the Gripen, Volvo Flygmotor worked with the US company General Electric to adapt that company’s F404J engine. Ericsson Microwave Division developed the Gripen’s PS-05/A radar, which is substantially more capable than the preceding PS-46/A radar fitted to the Viggen. To enable the PS-05/A radar to perform the multiple missions envisaged for the Gripen, Ericsson Microwave Division exploited its ERIEYE experience in incorporating active electronically scanned array technology. The Gripen’s multi-role capabilities and the greatly enhanced capacity of its radar and other sensors approached the limits of the existing STRIL 90 system. The introduction of the ERIEYE radar broad area surveillance radar was expected to exacerbate this problem after 1997.
Accordingly, in October 1990, FMV awarded CELSIUS Tech (now SAAB) what was then one of Sweden’s largest single computer contracts to radically upgrade the STRIL 90 command and control system. This culminated in the development of STRIC as a fully automatic air surveillance and operations control system but based, like its predecessors, on air defence sectors. Each air defence sector has a Sector Operations Centre which receives radar data from both static and mobile Control and Reporting Centres which are fed in turn by high- and low-level air surveillance radars. In 2001 contracts were awarded to upgrade STRIC and improve its integration with the JAS39 Gripen aircraft and the ERIEYE system. In 2004, in response to changing Swedish defence priorities, contracts were awarded to adapt STRIC for
179 Ibid., pp. 250-251. 180 Annika Brandstrom, Coping with a Credibility Crisis – The Stockholm JAS Fighter Crash of 1993, Swedish National Defence College, Stockholm, 2003, p. 20. 181 Greg Goebel, The SAAB JAS 39 Gripen, http://www.vectorsite.net/avgripen.html, p. 2, accessed 17 February 2010. 82
operation with NATO’s Air Command and Control System.182 STRIC provides two-way data links between the ground control centres and other air defence assets, including the Korpen signals intelligence aircraft, the ERIEYE broad area surveillance system and the Gripen fighters. Both ERIEYE and fighter radars can be controlled from the command centre, permitting remote vectoring and silent intercept.183
To help keep the cost, schedule and technical risk inherent in developing the Gripen to acceptable levels, the Industry Group JAS proposed, and the Swedish customer accepted, much higher levels of imported content than had been acceptable at earlier stages of the Cold War. In the case of the radar, for example, this entailed Ericsson Microwave Division importing from the US the high-precision/high-performance micro-circuitry required in the new radar. This relaxation of the earlier emphasis on indigenous ‘make’ solutions left the Swedish defence competence bloc more exposed to technological coercion than hitherto. Around 1981, for example, the US government learnt that a Swedish company, DataSAAB, had systematically breached COCOM guidelines and US export licence conditions in supplying strategically sensitive air traffic control radar technology to the USSR. The US informed the Swedish government and stopped processing export licences for Swedish purchase of, for example, infrared Sidewinder air-to-air missiles and licences to manufacture the GE 404 jet engine for the JAS39 Gripen fighter aircraft.
The US government demanded, and the Swedish government gave, a public apology. Ericsson, as the owners of Datasaab, and the Swedish government paid a $US3.125 million fine imposed by a US court in 1984. In the meantime, in May 1982, the Swedish government and the Federation of Swedish Industries introduced new more stringent arrangements for control and protection of dual use technology in Sweden. In deference to Sweden’s neutrality policy, the Swedish government stopped short of joining COCOM but secured US government agreement that the May 1982 arrangements provided equivalent protection. The US then lifted its embargo on defence materiel exports to Sweden.184
4.5 Swedish demand In executing demand for novel solutions to military requirements, the actors performing the customer function in the Swedish defence competence bloc will search for, select and procure those solutions they judge, ex ante, will provide the best value for money. Those judgements will have regard to relative economy (spending less on inputs), relative efficiency (output relative to input) and relative effectiveness (the impact achieved). This section analyses how Swedish institutions, military doctrine and technology influenced the
182 Jane’s C4I systems, 28 July 2011: STRIL/STRIC air defence system (Sweden), available at http://jc4i.janes.com/docs/jc4i/search/results, accessed 11 August 2011. 183 Jane’s C4I Systems, 5 July 2011: TARAS Swedish Air Force communication system, available at http://jc4i.janes.com/docs/jc4i/search/results, accessed 11 August 2011. 184 Inquiry on Security Policy: Peace and Security: Swedish Security Policy 1969-1989. Abridged Version and Translation of SOU 2002:108, Statens Offentliga Utredningar, Statsredsberedningen, Stockholm, 2004, pp 109- 113. 83
way the Swedish customer exercised demand for novel solutions to military capability requirements during the Cold War.
During the Cold War, the search by the Swedish defence customer for a combination of artefact and supplier that constituted acceptable value for money in meeting requirements was path dependent. The outcome of previous choices by the customer constituted the starting point for the next cycle of search by the customer for solutions to new requirements. Such path dependency of the search process was a product, firstly, of the dynamic stability of the SwAF’s ground-based air defence doctrine and, secondly, of the knowledge and competencies developed by Swedish innovators, entrepreneurs and industrialists in responding to previous cycles of demand.
The Swedish customer’s demand for politically neutral technology was elastic and the value placed by that customer on Swedish ‘make’ solutions to requirements for military capability had limits. Hence, in searching for an economical combination of artefact and supplier, the Swedish defence customer tested local design, development and production of candidate artefacts against imported options. For example, in searching for a solution to the SwAF requirement for a rapid reaction surveillance capability, FMV capability entrepreneurs did consider the suitability of overseas alternatives like the E2C Hawkeye and the E3 Sentry but ruled them out because they offered relatively poor value for money. As the development of ERIEYE illustrates, however, the detailed trade-offs required to identify the most economical solution at the system, sub-system and component level were made by Swedish industrialists working closely with an informed customer. Hence, for example, it was Ericsson Microwave Division that judged the appropriate combination of indigenous Swedish and imported US technologies in developing and producing the ERIEYE radar system, albeit in close consultation with FMV.
In searching for a combination of artefact and supplier, the Swedish defence customer also took into account the Swedish supplier’s previous performance in designing, developing and producing artefacts consistent with SwAF requirements. Swedish corporatism and the tendency for Swedish companies to specialise in particular technological systems meant that Swedish industrialists were efficient designers, developers and producers of path- dependent solutions to path-dependent requirements. Hence, for example, the Swedish customer confined its search for a solution to the SwAF’s requirement for a rapid reaction surveillance capability to Ericsson Microwave Division because of that organisation’s proven capacity to design, develop and produce successive generations of Swedish airborne radars.
Finally, in searching for an effective combination of artefact and supplier, the Swedish defence customer took into account two dimensions of effectiveness. The first dimension of effectiveness related to the anticipated military utility of the artefact proposed by the industrialist once the military user had accepted it into service. For example, by deploying the airborne ERIEYE radar, the SwAF expected to be able to detect, identify and classify in- 84
coming Soviet aircraft earlier and with more precision than was possible with ground-based surveillance radars. The second dimension of effectiveness related to the enhanced capability expected as a result of the military user incorporating the artefact proposed by the industrialist into that user’s socio-technical regime. For example, the SwAF envisaged embedding the ERIEYE radar into its ground-based air defence system by linking it to an upgraded STRIC command and control system and to the new Gripen multirole fighter. By embedding the ERIEYE system into its overall portfolio of air defence assets, the SwAF expected the enhanced portfolio to enable it to achieve the same or better probability of intercepting Soviet aircraft with fewer expensive Gripen fighters.
Selecting a solution to a Swedish capability requirement entailed the Swedish customer selecting both an artefact and a supplier of that artefact. During the Cold War, the Swedish customer demanded path-dependent artefact/supplier combinations. This pattern of demand and the supplier’s provision of path-dependent solutions to well-understood requirements enabled both customer and supplier to make well-informed assessments of the relative economy, the relative efficiency and the relative effectiveness with which proposed artefacts met the requirement. For example, FMV capability entrepreneurs were required to gauge the economy, efficiency and effectiveness with which the ERIEYE radar proposed by Ericsson Microwave Division could meet the SwAF requirement for a rapid reaction surveillance capability. In doing so, those capability entrepreneurs drew on extensive prior experience with Ericsson Microwave Division’s ability to estimate costs and detailed familiarity with the cost of the PS-46A radar. By building an ERIEYE technology demonstrator, both FMV and Ericsson gained a detailed understanding of the cost involved in combining the PS-46A radar with electronic scanned array technology to meet the SwAF requirement.
In procuring selected artefacts from selected suppliers, the Swedish defence customer was able to exploit the dense networks linking the actors involved in developing technologically advanced solutions to military capability requirements. Such networks enhanced the economy, efficiency and effectiveness of Swedish defence procurement by enabling the actors populating the defence competence bloc to exchange large volumes of tacit and uncodified information and knowledge efficiently and effectively. For example, dense networks between FMV capability entrepreneurs and Ericsson Microwave Division engineers enabled the latter to design, develop and produce six ERIEYE systems for the SwAF within six years of the decision to procure them. Similarly, Swedish corporatist norms enhanced the efficiency of Swedish defence procurement by encouraging industrialists to husband the knowledge they gained through prior learning and encouraging the customer to develop detailed knowledge of supplier capacity. For example, FMV’s detailed understanding of both ERIEYE artefact technology and Ericsson Microwave Division capacity coupled with the latter’s detailed understanding of the requirement enabled both parties to conclude a fixed price contract for the supply of six ERIEYE systems to the SwAF.
85
Dense networks between the Swedish customer and Swedish industrialists also enhanced the effectiveness of Swedish defence procurement by enabling both customer and supplier to make informed judgements about the associated cost, schedule and technical risk. Such judgements were informed by the Swedish customer’s willingness to accept and fund development risk and the Swedish industrialist supplier’s willingness to accept production risk. Such informed assessment of procurement risk enabled the Swedish customer and the Swedish industrialist to improve the effectiveness of Swedish procurement by making more informed trade-off between, on one hand, the cost, schedule and technical risk inherent in producing an artefact with enhanced military utility and, on the other hand, the enhanced capability value generated by successfully embedding that artefact in the appropriate military socio-technical regime. For example, in developing ERIEYE, FMV capability entrepreneurs and Ericsson Microwave Division engineers were able to reconcile divergent perspectives as to the appropriate trade-offs between cost and capability so as to meet not only SwAF requirements but also those of other countries.
Section 4.6 Conclusion This chapter has described the structure of the Swedish defence sectoral innovation system in terms of its constituent building blocks. The next step is to analyse the conduct of the Swedish defence sectoral innovation system by describing the development, production and procurement of the ERIEYE radar system. This is the subject of Chapter 5.
86
Chapter 5: Developing, Procuring, Operating and Diffusing the Erieye Radar System As the next step in analysing the performance of the Swedish defence sectoral innovation system, this chapter describes how it works. It does so by describing how Sweden’s domestic radar supplier, the Microwave Radar Division of L.M. Ericsson, developed the Erieye radar as a solution to the SwAF requirement for a rapid reaction radar system, how FMV procured the system developed by Ericsson, how the SwAF operated the system procured on its behalf by FMV and how the Erieye technology subsequently diffused. The chapter begins with an overview of this process. It then discusses how L.M. Ericsson developed the radar-related competencies required to design, develop and produce a rapid reaction surveillance radar that met the SwAF requirement. This is followed by a description of the requirement for a rapid reaction surveillance capability that prompted the development of Erieye. A discussion of the action taken to execute demand for a solution to that requirement then follows. The chapter concludes with a discussion of the diffusion of Erieye after its acceptance into SwAF service.
5.1 Developing Erieye – Overview Figure 5.1 provides an overview of the Erieye development process. Each of the steps illustrated in Figure 5.1 is described in more detail in the chapter. To summarise: firstly, FMV’s search for an economical solution to the SwAF requirement prompted Ericsson to synthesise an established technology (the PS-46A pulse Doppler radar) and a relatively novel technology involving beam forming by the electronic scanning or phasing of radar signals. Secondly, FMV’s selection and procurement of Erieye from Ericsson straddled the end of the Cold War which reduced the SwAF’s incentive to sponsor further Erieye development. Thirdly, Ericsson identified non-Swedish customers for Erieye whose demand provided the impetus for Erieye development after its introduction into SwAF service.
87
Figure 5.1 Erieye development overview
5.2: Developing radar competencies at Ericsson L.M. Ericsson was founded in 1876 to manufacture and repair telegraph equipment. By 1900 the company was selling telephones in both Sweden and abroad. As an internationally competitive, large-scale manufacturer Ericsson helped develop Sweden’s engineering industry.185 During World War Two Ericsson manufactured air defence and air raid warning systems, aeronautical and measuring instruments, telecommunications equipment and machine guns for the Swedish defence effort.186 Ericsson management also contributed to the work of the National Industry Commission in managing the acute supply problems that beset Swedish industry during the Second World War.187
During World War Two Ericsson reportedly worked with the Swedish military research agencies on ‘echo radio’, a precursor of radar. However, Ericsson’s post World War Two involvement in radar technology began with a contract, at the end of the war, to manufacture under licence a French search radar for the Swedish army. Ericsson subsequently produced under licence both tracking radar for anti-aircraft use and aircraft radar for the Swedish A-32 Lansen aircraft.188
Ericsson took less than 10 years to transition from producing imported radar designs under licence to designing, developing and producing radars for the Swedish military. This rapid transition was facilitated by the company’s extensive electronics-related research and development capacity which pre-dated its entry into the military radar business. Ericsson
185 John Meurling and Richard Jeans, The Ericsson Chronicle, Informationsforlaget, Stockholm, 2000, pp. 19- 49. 186 ibid., p. 169. 187 ibid, p. 172. 188 ibid., p. 252. 88
had established this capacity to support its exports of leading-edge telephone switching and exchange equipment in the face of stiff competition from, inter alia, US and British suppliers. Producing French radar designs under licence helped Ericsson adapt its established competencies in telephony to the new radar technology.
This adaptation was facilitated by the company’s creation of the Ericsson Microwave Division dedicated to developing radar-related competencies. L.M. Ericsson management deliberately chose to locate the Microwave Division close to Chalmers University in Gothenburg to take advantage of the pool of young electronics engineers graduating from that university. Thereafter, a combination of geographical isolation and a focus on research- intensive radar developmental work for the Swedish military led the Division to develop a distinctive ethos within the Ericsson company.189
Once established, Ericsson Microwave Division steadily enhanced its competence in a diverse range of radar technologies through a process of cumulative learning-by-doing sustained over some 50 years. The Division was encouraged to accumulate knowledge of radar-related technology by Sweden’s corporatist norms, the high value consistently accorded to indigenous ‘make’ solutions to radar-related requirements and the tendency for Swedish industrialists to specialise in specific technological systems. As a result, the organisation of skilled individuals and competent teams that began as L.M. Ericsson’s Microwave Division (and in 2006 became SAAB’s Electronic Defence Systems) remained Sweden’s prime repository of expertise in the development and production of radar during and after the Cold War.
During (and after) the Cold War, the individual radar-related skills and organisational radar- related competencies husbanded by Ericsson Microwave Division were refreshed and extended by Swedish procurement of increasingly sophisticated radars for land, maritime and airborne applications. Particularly important for present purposes were the cycles of Swedish air defence system upgrades defined by the development and introduction of successive generations of fighter aircraft. This regular refreshment of deep competencies worked to limit the gap between the Microwave Division’s competencies extant at a given time and those required to meet the informed Swedish customer’s demand for novel solutions to radar-related requirements at that time.
This cycle of competence refreshment and consolidation by periodic procurements followed by competence development to meet new but path-dependent demands by an informed customer enabled the Microwave Division to meet that new demand more quickly and cheaply than would otherwise have been the case. In addition, the demand by an informed military customer for militarily competitive solutions to radar-related requirements meant that the Microwave Division’s competencies were readily adapted to meet comparable
189 ibid., p. 256. 89
demands by non-Swedish customers. This facilitated the relatively rapid diffusion among non-Swedish customers of radar-related solutions to SwAF requirements developed by the Microwave Division.
The Ericsson Microwave Division developed radar-based competencies in both ground- based and airborne radar systems. This enabled considerable cross fertilisation across both activities. Because ground-based radar was less constrained by weight and volume considerations, Ericsson Microwave Division’s ground-based radar applications tended to lead its development work and to serve as a foundation for subsequent development of more constrained airborne radars.
By 1958 Ericsson Microwave Division had developed sufficient competence in radar technology to design, develop and produce a series of ground-based fire control radars (PE- 452, PE-48, AND PE-453) linked to anti-aircraft gun and surface-to-air missile components of Sweden’s air defence system. Ericsson Microwave Division also supplied radar components for naval fire control systems (in cooperation with Marconi Radar Systems Ltd of the UK). In 1970 Ericsson began developing the distinctive Giraffe family of ground-based search radars to meet a niche Swedish air defence requirement. As the name suggests, this radar had a distinctive folding mast which, when deployed, allowed its pulse Doppler search radar to see over nearby terrain features like trees, thereby extending the range at which it could detect low-altitude, narrow cross-section aircraft targets in an environment characterised by severe clutter and active electronic countermeasures.
The Giraffe system was first displayed in 1975. Ericsson Microwave Division designed it to complement the fixed ground-based radar elements of the STRIL command and control system by providing a mobile radar capacity able to fill gaps in radar coverage for air defence purposes. To this end the Giraffe system was mounted on an all-terrain vehicle and linked to Sweden’s RBS-70 air defence missile system. Ericsson Microwave Division (and, after the Division’s sale, SAAB Electronic System Division) continually upgraded the Giraffe system which now includes an integral command, control and communication capability.190 The system comes in several variants which provide a militarily competitive solution to a niche demand, having been sold to over 20 countries, including the US and Australia. In 1987 Ericsson Microwave Division extended its portfolio of ground-based radar programs by initiating development of the innovative ARTHUR artillery locating radar. The primary function of the ARTHUR system is to detect hostile artillery by detecting shells or mortar bombs in the up-going phase of their trajectory after firing, tracking them in flight, calculating their point of origin and impact and displaying the information to ARTHUR system operators. Ericsson Microwave Division (and, after the Division’s sale in 2006, SAAB
190 ibid., p. 254. 90
Electronic System Division) continuously upgraded the system which provides a militarily competitive solution to a niche demand, having been sold or leased to some 12 countries.191
Ericsson Microwave Division’s substantive involvement with airborne microwave radar technology began with production under licence of French CSF radars in the late 1950s. According to Meurling and Jeans, Ericsson began developing the first all-Swedish airborne radar (for the D version of the F-35 Draken aircraft) in close cooperation with SAAB and the SwAF in the second half of the 1950s. This pulsed radar (designated PS-03/A) was delivered 1961-65. This development work entailed minimal commercial risk to Ericsson because, in accordance with Swedish practice, the SwAF and FMV funded the work on a cost-plus basis. The competence Ericsson Microwave Division developed through the above licensed production rendered the cost, schedule and technical risk inherent in such arrangements acceptable to the SwAF customer. Thereafter, Swedish airborne radar development followed a clear technological trajectory. As Figure 5.2 indicates, Ericsson was able to sustain the trajectory by overlapping successive programs of radar development and production to meet the progressively more demanding requirements of successive generations of Swedish fighter aircraft developed to meet Swedish air defence requirements. Taken together with Ericsson’s ground-based radar programs, these overlapping radar development and production programs enabled the company to apply the technological and managerial lessons learnt in previous programs to the next generation of system.
Figure 5.2 Ericsson/SAAB: selected airborne radar development and production192
191 SAAB, ARTHUR Weapon Locating System available at http://www.saabgroup.com/Global/Documents%20and%20Images/Land/ISTAR/ARTHUR/ARTHUR%20ENG %20print.pdf, accessed 29 August 2011. 192 Adapted from SAAB: PS-05/A Advanced Aviation Multi-mode Radar, available at http://www.saabgroup.com/Global/Documents%20and%20Images/Air/Sensor%20Systems/PS%2005_A/PS05_ 100422.pdf, accessed, 28 August 2011. 91
Sustained production of previously developed radar systems also generated the financial flows required to fund the company’s contribution to development of the next generation of radar. During the Cold War, Ericsson Microwave Division accumulated radar-related know-what, know-why and know-how types of knowledge in the course of moving along the radar technology trajectory. This accumulated technical and managerial knowledge combined with Sweden’s policy preference for indigenous ‘make’ solutions to constitute a formidable barrier to entry to the Swedish radar market. That barrier persisted as long as the Division succeeded in meeting an informed customer’s exacting requirements and the perceptions of the Soviet threat remained. When these policy parameters changed in response to the collapse of the Soviet Union, the Division began engaging in more collaborative radar ventures (initially with GEC Marconi).
During the Cold War, the successive cycles of investment in Sweden’s ground-based air defence capability initiated by the Swedish defence customer prompted innovators in Ericsson Microwave Division to refresh their ability to devise novel combinations of old and new technologies. For example, the Division’s engineers used sub-miniature vacuum tube technology in the early, pulsed airborne radars developed for the SwAF. The Division began using semiconductor technology in airborne radars equipment in developing evolved radars for the F version of the Draken aircraft in 1964-70. The Division was able to take advantage of the semiconductor-related learning in this program when it participated in the next cycle of development of the Swedish air defence system.
The SAAB 37 Viggen aircraft was developed during 1962-72. FMV procured over 300 Viggen aircraft in several variants during 1971-90. Ericsson was responsible for development and production of the airborne target sensing equipment for the Viggen aircraft. Adapting the Viggen for the interceptor/air superiority role drove Swedish radar-related innovation, ratcheted up Ericsson Microwave Division competencies and established the technological foundation for the development of Erieye. This process of demand-driven learning-by-doing is described in the following paragraphs.
5.3 Developing a rapid reaction surveillance radar In developing a novel solution to the SwAF requirement for a rapid reaction surveillance radar, Ericsson Microwave Division drew on two streams of radar-related technology. One stream was pulse Doppler radar technology which the Division mastered in the process of adapting the Viggen aircraft for the interceptor/air superiority role previously performed by the Draken aeroplane. The second stream was electronically scanned array technology with which the Division had been experimenting for some years before deciding to apply it to the development of a rapid reaction surveillance radar for the SwAF. The process by which Ericsson Microwave Division developed its competence in these two radar technology sub- systems is described in the following paragraphs.
92
The upgrade of the SAAB AJ 37 Viggen’s avionics, including its radar, was driven by a change in Soviet airborne assault tactics in the 1960s. Hitherto, Soviet tactics had focused on attack at high altitudes. As the NATO countries deployed more capable ground-based radar in defence, the Soviet Air Force changed to attack at low altitudes, making use of the terrain to evade detection. By adopting low-level airborne assault tactics, the Soviet Air Force forced the defending aircraft to fly higher which increased their vulnerability. Such tactics also reduced the effectiveness of the conventional pulse search radars fitted to the defending aircraft at the time. In looking down to detect low-flying Soviet aircraft, such conventional radars had difficulty distinguishing the radar signal reflected by the Soviet aircraft from that reflected by the ground. This tactical disadvantage prompted NATO (and Sweden) to develop a ‘look down/shoot down’ radar.
To meet the SwAF requirement for a look-down/shoot down capability, FMV’s predecessors began working with Ericsson and other Swedish electronics companies to develop a pulse Doppler radar system. In order to distinguish a moving target from the stationary background, a pulse Doppler radar takes advantage of small shifts in the frequency of the return signal caused by the movement of the target relative to the background. The technology required to detect such small Doppler shifts was well beyond that required for conventional radars. For example, the transmitted signal had to be specturally ‘pure’ and the receiver had to be extremely stable. Any extraneous modulation of the signal (by, for example power supply units) had to be eliminated. Frequency filtering had to be adjustable to stop jamming.
Although these were essentially path-dependent developments of established radar technology, the knowledge and skills involved were significantly more advanced than those required to design, develop and produce conventional pulse radars. To meet the SwAF requirement, innovators in Ericsson Microwave Division had to develop the pulse Doppler radar technology, harden it to defeat Soviet electronic counter measures and miniaturise the signal and data processing system to fit into the limited amount of space available in the Viggen aircraft. This required Ericsson engineers to develop new competencies, including, for example, the design and development of multi-layer printed circuit-board assemblies.
The Division and its collaborators developed their first experimental pulse Doppler radar in 1963-65 and a second experimental version in 1965-67. These experimental radars were followed by three prototype pulse Doppler radars (designated PS-46A) for the interceptor/air superiority variant of the Viggen (designated the SAAB JA 37), developed during 1970-78. Ericsson Microwave Division delivered its first series produced PS-46A radar system on 26 April 1978, some 15 years after the initial pulse Doppler work – see Table 5.1.193 A key PS-46A innovation was the linking of the pulse Doppler radar to a digital signal
193 B. Andersson, J Nilsson and S-B Thelander, Radar PS-46A for Aircraft JA-37, Ericsson Review, Vol 60 (2), 1983 pp. 46-57. 93
processor and in turn to the central digital computer at the heart of the Viggen’s entire avionics system. Ericsson Microwave Division’s competence in digital technology, which was critical to this innovation, was boosted by Ericsson winning a large US Army order in the mid-1980s for digital communications technology. This key order for the MiniLink digital microwave system (which became one of Ericsson’s core products) cemented the Division’s competence in this emerging technology.194 The time Ericsson Microwave Division took to develop pulse Doppler technology and incorporate it into the PS-46A radar is summarised in Table 5.1 below.
The non-availability of cost data in Table 5.1 needs explanation: the research undertaken for this thesis uncovered valuable technical and project management-related information about the PS-46A radar published by the Division’s engineers, notably in the company journal Ericsson Review and in the proceedings of the Institute of Electrical and Electronic Engineers (IEEE). The research was unable to find reliable data on the cost incurred by the Division in developing the PS-46A radar. For this thesis, however, it is important to note that the cost of developing the PS-46A radar was attributed to the Viggen upgrade program. As a sunk cost, it was not attributed to the ERIEYE program and can therefor be ignored in this analysis of the performance of the Swedish innovation system.
Table 5.1 Developing Ericsson competency in pulse Doppler radar technology
Dates Activity Time elapsed (years) Estimated Cost (SEKM)
1963-65 First experimental 2 Not available pulse Doppler radar
1965-67 Second experimental 2 Not available pulse Doppler radar
1970-78 Prototype PS-46A 8 Not available radar
1978-87 Series production of 9 Not available PS-46A radar
Total 21 Not available
As already indicated, the pulse Doppler technology incorporated in the PS-46A radar was one of two radar technology sub-systems exploited by Ericsson Microwave Division in developing the ERIEYE radar. The second radar technology sub-system was electronically
194 Meurling and Jeans, p. 263. 94
scanned or phased array technology. A phased array is an array of antennas in which the relative phases of the electronic signals fed into the antenna are varied in such a way that the radio waves transmitted by the antenna are reinforced in a desired direction and configuration while being suppressed in undesirable directions and configurations. A phased array antenna can generate wide beams for searching, narrow beams for tracking, flat fan- shaped beams for height finding and narrow pencil beams for terrain following. In a hostile jamming environment phased array technology can be used to effectively block a jammer from entering the receiver chain. While phased arrays have important limitations, one of the technology’s key advantages in surveillance applications is that it obviates the need to mechanically point the antenna in the direction of the target. The US Department of Defense began sponsoring research into phased array technology at Lincoln Laboratory (part of the Massachusetts Institute of Technology) in 1958.
Gallium arsenide-based micro-electronics were a key enabling technology in the development of phased arrays for military purposes. The technology was pioneered by the Lincoln Laboratories in the late 1960s as part of its work on phased arrays. Gallium arsenide- based devices can operate at extremely high speeds and frequencies, not matched by silicon-based micro-electronics used in commercial applications. Gallium arsenide micro- electronics are used to generate the very high-frequency/short wavelengths used to form narrow, high-resolution radar beams for tracking small, fleeting military targets like cruise missiles and high-performance fighters. The US released phased array technology and the associated gallium arsenide based microcircuitry to Sweden.
The first generation of phased array technology used essentially conventional radar architecture. In this so-called passive array technology the antenna changed but the rest of the radar system, including signal processing, remained unchanged. Early passive array technology required additional computer hardware to control the antenna shifters, all of which added to weight and required extra electrical power which had to be supplied by larger accessory generators. These disadvantages were largely eliminated by the introduction of gallium arsenide micro-electronics in the 1980s. This enabled dramatic reductions in the weight and volume of the electronics required for both radar transmitter and receiver. For example, an entire radar (transmitter, receiver and antenna) could be built as a single transmitter-receiver module about the size of a house brick. The gallium arsenide micro-electronics enabled the introduction of active phased arrays and, in turn, application of the technology in airborne early warning and control systems.
In active phased array antennas, each array element or group of elements has its own miniature microwave transmitter, obviating the need for the single large transmitter tube of the older passive array technology. Each element comprises a module which contains the antenna slot, phase shifter, transmitter and, often, a receiver. Because different modules can operate at different frequencies, active phased arrays can produce numerous sub-
95
beams and actively illuminate a much larger number of targets than the earlier passive phased arrays. The gallium arsenide-based micro-electronics used in solid state transmitter- receiver modules enable the latter to broadcast effectively at a much wider range of frequencies, enabling the active phased array radar to change frequency with every pulse transmitted. This enhances the rapid scanning, adaptable beam-forming and rapid steering features of phased array radars.
Ericsson Microwave Division invested in phased array technology because a single antenna incorporating phased array technology can perform, almost simultaneously, a range of functions that would require several purpose-built antennas using conventional technology. Ericsson began working with Chalmers University researching and developing electronically scanned array technology in 1968 and, by 1971, had constructed an experimental radar which used a passive phased array to track four targets simultaneously.195 These encouraging results led Ericsson and Chalmers to develop in the 1970s two additional technology demonstrators. One demonstrator was designed to test the feasibility of using electronically scanned arrays for electronic countermeasures, the other to test the feasibility of using the technology for satellite detection and tracking.
In the 1980s Ericsson used the satellite detection demonstrator as the basis for its development of active electrically scanned arrays for battlefield surveillance radars and for airborne early warning. In the latter application the technology was configured to provide a large azimuth scanning sector, very low sidelobes, high pointing accuracy and low weight. Ericsson began series production of active electronically scanned arrays for the battlefield surveillance radar in 1992, slightly ahead of its production of such arrays for airborne early warning applications. The latter began in 1993.196 The time taken by Ericsson Microwave Division and FMV in developing the electronically scanned array technology and in incorporating that technology in the Erieye rapid reaction surveillance radar is summarised in Table 5.2 below.
195 O. Dahlsjo, Antenna research and development at Ericsson, IEEE Antenna and Propagation Magazine, Volume 34 (2), April 1992, pp. 7-17. 196 ibid. 96
Table 5.2 Developing Ericsson competency in electronically scanned array technology
Dates Activity Time Elapsed (years) Estimated Cost (SEKM)
1963-71 Experimental passive 3 Not available phased array
1972-80 Passive phased array 8 Not available technology demonstrator
1980-90 Prototype active 10 Not available phased array
1993-2103 Series production of 20 Not available for active phased arrays for Erieye (SwAF & export)
Total 41 Not available
The non-availability of cost data in Table 5.2 needs explanation: search for reliable data on the cost incurred by Ericsson in developing electronically scanned arrays encountered the same difficulties as the search for data on the cost of developing the PS-46A radar discussed above. In 1995, however, the FMV reportedly paid Ericsson SEK 44 million for a multi-year study of active phased array technology for an upgrade of, among other systems, the JAS39 Gripen radar.197 This is equivalent to about 4% of the SEK 1200 million FMV paid for the six SwAF Erieye radars in 1992. Because Ericsson spread the cost it incurred in its early development of electronically scanned array technology over several radar development programs (of which Erieye was only one), it is reasonable to conclude that the cost of the electronically scanned array component of Erieye development did not drive the overall costs of designing, developing and producing the six Erieye systems for the SwAF and, therefore, can be ignored in this assessment of the performance of the Swedish innovation system.
5.4: The requirement for a rapid reaction surveillance system The look down/shoot down capabilities of the PS-46A radar enabled Viggen pilots to engage low-flying aircraft once they had intercepted them. The problem driving Sweden’s Cold War requirement for a rapid reaction surveillance capability, however, was obtaining sufficient warning of attack by Soviet aircraft to achieve timely intercept. Like British air defence
197 Jane’s International Defense Review, Sweden Studies Active-Array Radar, 1 January 1995. 97
planners in 1939, Swedish defence planners needed to obtain sufficient early warning of approaching hostile aircraft to enable Swedish fighters on strip alert to scramble from the ground and achieve intercept.198 Options like deploying intercepting aircraft on continuous combat air patrol were considered not only prohibitively expensive but also of dubious effectiveness. Developments in Soviet airborne assault capability also led Swedish air defence planners to seek more flexibility to enable effective responses to rapid changes in the level and direction of the Soviet threat. This entailed providing warning of attack from, say, Sweden’s south or south west and providing radar surveillance of otherwise uncovered directions within minutes.
During the 1950s-1970s, Sweden invested in a comprehensive national sensor grid, which included traditional ground-based radar of different types and associated command and control arrangements. This system provided adequate coverage of aircraft movements at high altitude. During the late 1960s improvements in ground-based radars and associated air defence systems caused the Soviets to change tactics and to adopt low-level strikes by aircraft flying at supersonic speeds and at altitudes of 600 ft or less. Ground-based radar could not provide sufficiently early warning of aircraft operating at these speeds and altitudes. For example, when teamed with interceptor aircraft on strip alert, ground-based radars along Sweden’s eastern border provided a detection line for low-level attacks of some 50 km from the Swedish border, an engagement decision line of some 40 km from the border but a defeat line inside the border. In tactical terms, this rendered the border line radars vulnerable to attack and, hence, compromised the viability of Sweden’s air defence.
In these circumstances, the overriding operational requirement driving demand for a rapid reaction surveillance capability was the need to extend the 50 km detection line and the 40 km decision line for engaging incoming aircraft and missiles provided by ground-based radar. Swedish air defence planners also sought to increase the resilience of the nation’s air defence system by protecting the ground-based radar system, the intercept aircraft and the bases from which they operated and the command and control system. To this end, the rapid reaction system also needed to provide a ‘gap filler’ replacement of Swedish ground- based radar coverage lost as a result of, for example, technical failures or battle damage. This required a system sufficiently responsive to provide for immediate re-establishment of lost coverage.
5.5: Executing demand for a rapid reaction surveillance system The SwAF formulated its requirement for an airborne early warning-based rapid reaction surveillance capability in 1967-68, before the technology required to produce an effective, survivable and affordable solution to that requirement existed. FMV monitored evolving airborne early warning technology and, as explained below, began to define a feasible
198 The following material is based on Carl-Gilbert Lonroth, interview, Stockholm, 16 November 2009. 98
system adapted to Sweden’s needs in the late 1970s.199 In 1978-80, Carl-Gilbert Lonroth, an ex-SwAF electronics engineer working in FMV’s Radar Division, commissioned Ericsson to investigate the feasibility of adapting the PS-46/A radar developed for the JA-37 Viggen aircraft to meet the Air Force’s surveillance radar requirement.
A key reason for the focus on the PS-46/A radar was the latter’s advanced, but proven, look down/shoot down capabilities based on its ability to handle variations in weather, its ability to distinguish targets from ground clutter and its resistance to electronic counter measures.200 According to Lonroth, the SwAF was only prepared to support development of a surveillance radar with a look-down detection range of a mere 60 km from the Swedish Baltic coast. In Lonroth’s view, however, detection of targets at anything less than 100 km range would simply not provide sufficient warning for effective intercept action. To detect targets at this range, however, the PS-46/A radar would require a larger and more powerful (and therefore more expensive) antenna.
The 1978-80 Ericsson studies suggested that a combination of PS-46/A radar and active phased array antenna technologies in a ‘pod’ mounted externally to an aircraft constituted a potential solution to the Air Force’s requirements at acceptable cost, schedule and technical risk. In 1985, in order to confirm the studies’ conclusions and refine assessment of the risk involved, FMV/Lonroth engaged Ericsson to develop a technology demonstrator based on their ‘pod’ concept. According to Lonroth, FMV paid Ericsson a fixed price (SEK 43 million) for the demonstrator. Ericsson contributed about the same amount because it envisaged developing an export version of the system if the demonstrator proved successful. 201 Lonroth/FMV specified a range for the demonstrator slightly longer than that stipulated by the SwAF, relying on Ericsson’s incentive to meet, at no extra cost to FMV, the more demanding range requirements of export customers. In the event, Ericsson’s demonstrator showed that the combination of PS-46/A radar and active phased array technology was able to detect targets far beyond the specified ranges, with the actual capability dependent on cost performance trade-offs. Earlier Ericsson studies had established the military utility of an adapted PS-46/A radar linked to an active phased array antenna housed in a ‘pod’ mounted externally to an aircraft. To sustain further development of the innovation, however, the ‘pod’ had to be married to a suitable platform and then integrated into the wider air defence system to obtain sufficient military capability value to justify the investment.
To inform the selection process, Lonroth once again engaged Ericsson Microwave Division, this time to investigate the feasibility of mounting a phased array radar under a SwAF fighter. Ericsson’s studies indicated that a fighter-mounted phased array radar had compelling disadvantages: such a platform/system combination could only detect targets at
199 Jane’s International Defense Review, Leading-edge technology for Swedish AEW, 1 March 1988, available at https://janes.ihs.com/CustomPages, accessed 6 August 2014. 200 Carl-Gilbert Lonroth and Sven Larsson, interview, 16 November 2009, p. 10. 201 ibid. 99
ranges of less than 100 km. A fighter could carry the phased array radar or weapons but not both. Deployment of a fighter-mounted phased array radar was constrained by a fighter’s limited loiter capacity. Mounting a phased array radar on a fighter prevented the fighter changing roles, thereby reducing the commander’s tactical options. Finally, such an arrangement confined the fighter to sub-sonic speeds, thereby forcing the fighter pilot to jettison a very expensive radar pod if attacked. In order to obtain the platform-related knowledge to interpret Ericsson’s findings, Lonroth liaised with LTCOL Sven-Olov Westerlund, a SwAF officer posted to the aircraft directorate of FMV. Westerlund advised that, having regard to the results of the Ericsson study, the tactical value gained by deploying a fighter-mounted phased array radar was far outweighed by the opportunity cost to the SwAF’s air defence capability inherent in the resulting degradation of fighter performance. Westerlund went on to suggest, however, that a turbo-prop aircraft would provide the requisite endurance and have the capacity to transport a phased array radar with worthwhile performance (in excess of 100 km).202
In considering the idea of mounting a pod on turbo-prop aircraft, both Lonroth and Westerlund envisaged mounting it underneath the aircraft fuselage. This led them to identify the Fairchild Metro turbo-prop aircraft (used by SAAB as a corporate aircraft) as a potentially appropriate platform for trialling the idea of a pod-mounted phased array radar. They were initially attracted to the Fairchild solution because the aircraft’s air intakes were mounted on top of the aircraft’s two engines and, hence, the aircraft’s propellers were lower to the ground. This required the aircraft to use a longer undercarriage which would enable it to accommodate the pod underneath the fuselage. Both Lonroth and Westerlund were concerned, however, that mounting a 9 m long pod underneath the aircraft’s fuselage would greatly restrict its capacity to manoeuvre on take-off and landing. This led Lonroth to consider emulating NASA’s transport of space shuttles on back of a Boeing 747 and fitting the pod on the top of the aircraft’s fuselage. As such dorsal mounting of the pod involved aeronautical engineering issues beyond Lonroth’s competence, he once again approached Westerlund. The latter had the organisational legitimacy to commission such a study and the aeronautical expertise to interpret the results.
Westerlund agreed to sponsor the requisite wind tunnel tests by Fairchild. This US company confirmed the basic feasibility of mounting the pod on the back of a turbo prop aircraft. Lonroth then focused on developing an electronically controlled phased array surveillance radar within the physical parameters set by what a turbo prop aircraft could carry on its back. But while Lonroth and Westerlund could envisage the way ahead for development of a Swedish airborne surveillance radar, they had to gain SwAF approval for the approach they proposed. Specifically, FMV funding of Lonroth’s airborne surveillance radar development was subject to SwAF endorsement.
202 Lonroth and Larsson, interview, 16 November 2009. 100
According to Lonroth, the SwAF was fully cognisant of the potential capability value of an airborne surveillance radar system but would not give FMV open-ended approval to develop that system. In effect, the Air Force judged that the greater the range at which the airborne surveillance radar system could detect targets, the more it was going to cost and the more it would divert resources from other Air Force priorities. Hence, the SwAF ignored Lonroth’s argument that the limited enhancement of the existing ground-based surveillance radar capacity it envisaged was simply not cost effective and refused to fund the development of any airborne surveillance radar system with a range in excess of 60 km from the Swedish coast. Lonroth then approached Ericsson Microwave Systems. He noted that FMV could fund the development work needed to meet the relatively undemanding and low-risk SwAF requirement of a range of 60 km from the Swedish coast. But Lonroth also sought to ensure that Ericsson did not undertake the developmental work in a way that ruled out detection at longer ranges.203 FMV IP policy provided the incentive for Ericsson to accommodate Lonroth’s ambitions within the FMV funding constraints. FMV policy was to allow Swedish companies to appropriate IP accruing through FMV development projects, subject to FMV retaining user rights. Lonroth was able to secure Ericsson’s agreement to accommodate the extra costs associated with achievement of longer range surveillance radar on the basis that Ericsson’s ownership of the resulting IP enabled the company to develop the radar for export.204
According to Lonroth, the Ericsson radar demonstrator indicated that a range of some 350 km was feasible. This far exceeded the FMV/Air Force requirement. This outstanding performance gave Lonroth sufficient confidence to reduce Erieye procurement time by authorising Ericsson to purchase long lead items in 1989, before concluding a contract with Ericsson for full-scale production of Erieye in 1993. More generally, the performance encouraged Ericsson to consider opportunities to export the system as soon as SwAF requirements were met. In taking the above action, Lonroth acted as a capability entrepreneur. In doing so he did not depart from the established pattern of Swedish military technological innovation being led by a well-informed customer on a demand-pull basis. Instead he worked within the customer element of the Swedish defence competence bloc to ensure that the SwAF’s short-term needs for a rapid reaction surveillance system were met without prejudicing future options for exploiting the system’s capabilities in the longer term.
Flight testing of the Ericsson demonstrator began in 1990, some five years after the initial design work for the demonstrator. Development of a viable demonstrator involved numerous trade-offs. These were facilitated by the dense networks between innovators and entrepreneurs located in both FMV and Ericsson. Innovators and entrepreneurs in Ericsson
203 ibid. 204 But see Jordi Molas-Gallart and P. Tang: Ownership matters: intellectual property, privatisation and innovation, Research Policy, Vol 35, 2007, especially pp. 209-211. 101
were entirely free to suggest cost/capability trade-offs to their counterparts in FMV. Trusted actors in both Ericsson and FMV were able to exchange information across dense networks in uncodified form, thereby reducing the time taken to settle the Erieye design. In parallel with the above FMV/Ericsson discussions, the FMV capability entrepreneurs had to reach agreement on who had responsibility for the integration of the radar into the SwAF’s evolving command and control system. Underlying these issues was the divergent requirements of, on one hand, Air Force operational doctrine and, on the other hand, FMV engineering logic. The issue focused on the location of the radar operator.
On one hand, and as a pilot-run organisation, the SwAF was inclined to locate the radar operator on board the aircraft platform. This meant removing responsibility for radar operation from the ground-based radar operators. On the other hand, Lonroth, Westerlund and other FMV engineers considered that putting a radar operator on board the platform would increase the complexity of the integration task and, hence, the cost and schedule risk involved. This engineering logic prevailed, so that both FMV as agent and the SwAF as principal agreed that surveillance radar system development would not provide for an on- board operator. With this issue settled, procurement activity then proceeded along two tracks. Within the FMV’s Aircraft Directorate, Westerlund assumed prime carriage of aircraft procurement and installation of the radar via a contract with SAAB. Lonroth, based in FMV’s radar directorate, retained responsibility for developing the radar and associated data links via a contract with Ericsson. In effect, these arrangements meant that FMV acted as prime contractor for the Erieye project, an arrangement made practicable by FMV’s combination of Erieye-related technical expertise, business acumen and deep knowledge of SAAB and Ericsson’s capacity as industrialists.
The Erieye demonstrator program established a firm basis upon which both FMV (acting as customer) and Ericsson (acting as industrialist) could assess the technical risk involved in developing Erieye as a solution to the SwAF’s rapid reaction surveillance requirement. A clear understanding of the risk involved was critical to the functioning of both capability and commercial entrepreneurs. For example, on the basis of the demonstrator program Lonroth judged that the risk inherent in developing Erieye was sufficiently low, and the demand for the capability sufficiently strong, to warrant his authorising procurement of long lead items and committing to long lead development in parallel, rather than subsequent to, other critical design and development activity. In particular, Lonroth felt sufficiently confident in the Erieye design to set system performance specifications before the Ericsson engineers had determined the systems output power (which determines the detection range). In doing so he structured the project specifications to retain the option of increasing the output power (and hence the detection range) before the design was frozen.
Importantly, none of these arrangements entailed extra cost for FMV. Ericsson actors performing the Erieye entrepreneur and industrialist functions were able and willing to
102
accommodate FMV demands because they understood the technical risk inherent in meeting the SwAF requirement and because meeting the SwAF requirement was the key to exploiting the system’s export potential. The SwAF accepted Erieye into service in 1997, a little over a decade after FMV and Ericsson decided to invest in the demonstrator.205 The following paragraphs describe the tactical adjustments required to realise Erieye’s military capability value in SwAF service and ongoing development of Erieye after its acceptance by the SwAF.
Ericsson designed Erieye around a two-sided active phased array, housed in a beam-shaped structure, carried above the fuselage of a twin-engine SAAB 340B commuter airframe. The SwAF Erieye arrangement could not provide 360-degree coverage, using conventional active phased array technology – a significant operational limitation. With each array scanning a 120-degree sector, the two-sided array had a 60-degree blind sector over the nose and the tail of the aircraft, and degraded antenna performance beyond 45 degrees off the beam of the aircraft. This was acceptable to the SwAF, given the nation’s geography and, during the Cold War, the axis of Soviet threat from the east. The SwAF compensated for the deficient coverage by deploying multiple ERIEYE units to cover a single area: for example, by operating in pairs, Erieye aircraft could patrol in two racetrack orbits set 90 degrees apart, thereby providing overlapping coverage. This sort of arrangement was entirely compatible with Sweden’s doctrine of ground-controlled intercept.
Critical to Erieye success in SwAF usage, however, was the capability of the computer data link and the STRIC networking which links the Erieye units to each other and to the ground air defence centre in generating a comprehensive picture of the air situation. In commissioning the development of Erieye, FMV envisaged operating it on the same principles used for Swedish ground-based radars, almost all of which could be controlled remotely and operated unattended. In Swedish service, radar signals are processed on board the aircraft and the resulting threat warning and other data is down-linked to a central ground control centre which handles aircraft and missile warning and intercept functions. As Figure 5.3 indicates, embedding Erieye into the Swedish air defence system greatly increased the warning the SwAF had of any Soviet air attack. More timely and accurate information about targets increased the economy of Swedish air defence by enabling the SwAF to achieve equivalent probability of intercept with fewer aircraft. It increased the efficiency of Swedish air defence by improving the probability of existing SwAF air combat aircraft being able to interdict sufficient attacking aircraft to render this an unattractive option for a would-be adversary. Finally, it enhanced Swedish security by helping the SwAF preserve the outer ‘shell’ of Sweden’s layered defence system.
205 Ibid., pp. 11-12. 103
Figure 5.3 Erieye coverage for airborne early warning206
Erieye deliveries to the SwAF began in 1993 (two years after the formal dissolution of the USSR in 1991) and were completed in 1999, by which time Swedish defence planners had accepted that the dissolution of the Soviet Union greatly reduced (but did not entirely eliminate) Sweden’s traditional concern about the security of its eastern approaches. In response, the SwAF substantially reduced the readiness of the SwAF units operating the Erieye system but stopped short of disbanding them.207 In accordance with Swedish support for post Cold War European security arrangements, the SwAF took steps to make the Erieye units interoperable with NATO air defence systems (including adapting them for the NATO standard Link 16 data transfer system). Finally, the SwAF and FMV provided active support for Ericsson’s efforts to export the system once it had completed SwAF deliveries. With SwAF requirements satisfied, the impetus for Erieye development shifted away from domestic Swedish demand to overseas demand and, hence, encouraged diffusion of the technology.
206 Jane’s Avionics, Erieye AEW&C Airborne Early Warning & Control Mission System Radar, available at http://www4.janes.com/subscribe, accessed 29 August 2011. 207 SwAF FSR890 operators, interview, Linkoping, 18 November 2009. 104
The time FMV and Ericsson Microwave Division took to develop Erieye and the cost they incurred up to delivery to the SwAF are summarised in Table 5.3 below.
Table 5.3 Developing and producing Erieye for the SwAF
Dates Activity Time Elapsed (Years) Estimated Cost (SEKM)
1978-80 Erieye feasibility 2 studies
1985-90 Erieye technology 5 43 demonstrator
1993-99 Erieye series 6 1200 production
Total 13 1243
5.6: Post-acceptance Erieye diffusion SAAB continued adapting the system to the needs of export customers which enabled it to continue developing the Erieye system. Brazil was the first international buyer. In 1998 the Brazilian government signed a contract for five radars, command and control systems and associated data link interfaces for use by the Brazilian Air Force to monitor the Amazon basin.208 The radars were installed on Brazilian Embraer EMB 145 aircraft, with Embraer being the lead system integrator. Deliveries were completed in December 2003. The Brazilian version of the Erieye radar incorporated improved software that extends the maximum sweep sectors on each side of the radar to 150 degrees (compared to 120 degrees for the SwAF version). In 1999 the Greek government signed a contract for four AEW&C aircraft with a joint venture between Ericsson Microwave Systems AB (Sweden) and Thales Airborne Systems (France) – called Ericsson-Thales AEW Systems (ETAS). Innovative business arrangements included meeting the Greek requirement for an interim capability by leasing SwAF aircraft/Erieye. Path-dependent development of the final version for Greece included modification of the Greek Erieye system for full integration into the NATO communication system. Final deliveries were completed in 2005.209 In June 2004 Mexico took delivery of one EMB 145 aircraft fitted with the Brazilian version of Erieye. In order to
208 See http://www.airforce-technology.com/projects/emb/ for more details. 209 Larsson, interview. 105
conform with Mexican requirements, Ericsson and Embraer modified the Brazilian version of the system with the addition of a satellite communications suite.
In 2006, after Ericsson Microwave Division was sold to SAAB, the Pakistan Air Force contracted with SAAB Surveillance Systems for the supply of five SAAB 2000 AEW&C aircraft. The latter is a variant of the SAAB 2000 regional transport turboprop aircraft but modified (by SAAB Aerotech) for the AEW&C role. The Pakistan Erieye design incorporated not only previous enhancements for other customers but also new-generation radar transmit/receive modules with 60% higher output to increase its range and enable it to detect hovering helicopters and track small naval targets to a range of 350 km.210
In November 2007, in conjunction with their purchase of Gripen fighter aircraft, the Thai government announced the purchase of two Erieye systems mounted on SAAB S100 Argus aircraft to meet their AEW&C requirements. In 2009 the United Arab Emirates government ordered two Erieye systems mounted on SAAB 340 aircraft. The time FMV and Ericsson Microwave Division/SAAB Defence Electronic Systems took to adapt Erieye for non-Swedish customers and the cost they incurred in doing so up to 2013 are summarised in Table 5.4.
The research on the pattern of post-acceptance diffusion of Erieye encountered the same problem relating to reliable information about the cost of post-acceptance Erieye systems. The Jane’s database, Ericsson/SAAB press releases and other sources do provide indicative information about the price paid by some non-Swedish buyers of Erieye. But, unsurprisingly given the commercial sensitivities involved, the research did not yield any information about the cost Ericsson/SAAB incurred in adapting the Erieye technology to suit the requirements of non-Swedish customers. Nevertheless, a broad indication of the expenditure incurred by Ericsson/SAAB in adapting Erieye for export can be inferred from its sale of six Erieye units to the SwAF for SEK 1,2 billion and assuming, say, a profit of 10% (in line with that of the earlier Framework Agreement). This suggests that Ericsson/SAAB might have spent in the order of SEK 3 billion in designing, developing and building 18 ERIEYE systems for non- Swedish customers – see Table 5.4 below.
210 See http://www.airforcetechnology.com./projects/saab-2000/ for more details. 106
Table 5.4 Adapting Erieye for export
Dates Activity Time Elapsed (years) Estimated Cost (SEKM)
1998-2003 Design & delivery of 5 5 Not available Erieye/Embraer EMB 145 for Brazil
1999-2005 Lease followed by 6 Not available design &delivery of 4 Erieye/Embraer EMB145 to Greece
2004 Mexico buys 1 1 Not available Erieye/Embraer EMB 145
2006-13 Design & delivery of 4 7 Not available Erieye/SAAB 2000 AEW&C aircraft to Pakistan
2007-2013 Design & delivery of 2 6 Not available Erieye/SAAB 2000 AEW&C aircraft to Thailand
2010-2011 UAE buys 2 1 Not available Erieye/SAAB 340 aircraft
Total 18 systems 13 years (1998-2013) Approx 3,000211
5.7 Conclusion The development of Erieye demonstrated how the various elements of the Swedish defence sectoral innovation system functioned in generating a novel solution to a Swedish requirement for military capability. The next step is to analyse how those elements affected the performance of the system in terms of the time taken to develop Erieye, the cost incurred in doing so and the pattern of Erieye’s development following its acceptance into SwAF service.
211 Estimate based on the sale of six ERIEYE systems to the SwAF for SEK 1.2 billion, implying a unit sale price of SEK 200 million, less a profit of 10% giving an indicative unit cost price of SEK 180 million for Erieye exports. 107
Chapter 6: The Swedish Defence Sectoral Innovation System and Erieye Innovation Outcomes This chapter analyses how the various elements of the Swedish defence sectoral innovation system influenced the performance of that system. To this end the chapter focuses on how those elements of the system influenced Erieye innovation outcomes in terms of the time taken to develop an Erieye-based solution to the SwAF requirement for a rapid reaction surveillance capability, the cost incurred in doing so and the trajectory of Erieye development after the system was accepted for SwAF service. The chapter is organised around the five building blocks of a defence sectoral innovation system: the chapter begins with a discussion of how Swedish armed neutrality, corporatism and governance affected Erieye innovation outcomes. This is followed by a discussion of how the performance of the various defence competence bloc functions by the actors populating the Swedish defence innovation system affected Erieye innovation outcomes. Then follows separate analyses of how those outcomes were affected by, respectively, Sweden’s doctrine of ground-based air defence and Sweden’s ground-based air defence technology base. The chapter concludes with a discussion of how Erieye innovation outcomes were affected by the way the Swedish customer searched for, selected and procured the system.
6.1: The influence of Swedish institutions For the duration of the Cold War Swedish armed neutrality fostered a durable consensus among Swedish defence actors in favour of ‘make’ solutions to, among other capability requirements, the SwAF requirement for a ground-based air defence capability. This consensus in support of ‘make’ solutions encompassed airborne radar. Successive cycles of airborne radar development gave all elements of the Swedish defence competence block an incentive and an opportunity to accumulate radar-related knowledge through learning-by- doing, learning-by-using and learning by the interaction of user and producer.
As part of this process Ericsson Microwave Division innovators accumulated deep knowledge of radar-related technology. Ericsson Microwave Division entrepreneurs and industrialists translated that knowledge into the competences required to design, develop and test an Erieye prototype. The Division’s radar-related competence was sufficient to produce an Erieye prototype that met the SwAF requirement within seven years of demonstrating (in 1980) the initial feasibility of using electronically scanned arrays to adapt the PS-46A radar for rapid reaction surveillance purposes. The learning accumulated also enabled Ericsson Microwave Division to design and produce a prototype within five years of a decision to do so (in 1985). This was accomplished at the modest cost of SEK 43 million, equivalent to about 20% of the estimated unit cost of the radars produced for the SwAF. 108
The process of designing, developing and producing Erieye for the SwAF enhanced Ericsson Microwave Division’s competence in applying the novel combination of pulse Doppler radar and electronic scanned array technologies to comparable requirements. Once the SwAF had accepted the Erieye system into service, Ericsson entrepreneurs then took advantage of the Ericsson Microwave Division’s enhanced competence when they marketed the system to overseas users as diverse as Brazil, Mexico, Greece, Thailand, Pakistan and the United Arab Emirates. Similarly, Ericsson innovators built on prior technological learning in adapting the system to the particular needs of these diverse customers while Ericsson industrialists harvested previously developed production know-how in producing competitively priced systems. (This thesis has estimated that Ericsson/SAAB could produce Erieye systems tailored to the export customer’s needs for in the order of SEK 180 million each). The depth and flexibility of Ericsson Microwave Division competencies can be gauged by the success of the Division’s innovators, entrepreneurs and industrialists in adapting the Erieye system to different customer requirements. These included different platforms (for example, SAAB 340, SAAB 2000 and Embraer aircraft), different airborne surveillance doctrines (for example, on-board data processing and ground-based data processing) and different operational environments (requiring, for example, NATO standard Link 16 or nation-specific data links).
During the Cold War a widespread consensus among Swedes about the threat posed by the Soviet Union helped foster the corporatism that became such a distinctive feature of the Swedish defence sectoral innovation system. In the case of Swedish ground-based air defence, corporatism denoted a relationship between the Swedish defence customer and a cluster of Swedish companies each specialised in a particular air defence technological system. During the Cold War, the Swedish government directed Swedish air defence business to these companies over successive cycles of air defence capability development, shared the risks they faced in developing novel solutions to Swedish air defence requirements and gave them rights to the IP they generated on its behalf. In return for these privileges, the companies invested in the innovator, entrepreneur and industrialist competencies required to generate novel solutions to Swedish military requirements.
The prevailing corporatist norms enabled Ericsson Microwave Division to develop over some five decades deep competences in rapidly evolving ground-based and airborne radar technologies. Such deep competence enabled Ericsson to transition rapidly from development of the Erieye technology demonstrator to complete production of the six radars procured for the SwAF: the latter task was completed within six years of the decision to proceed in 1993. Swedish corporatism helped Ericsson Microwave Division develop deep competence in a particular technological niche defined by pulse Doppler radar technology, electronically scanned array technology, commercial airline platforms and digitised data processing (both airborne and ground based). From this technological niche, established by meeting the Swedish customer’s informed but limited demand for six Erieye systems, the
109
innovators, entrepreneurs and industrialists fostered by Ericsson Microwave Division (and, after 2006, SAAB Defence Electronic Systems) delivered 18 systems to six countries over the 13-year period 1998-2011. This diverse range of non-Swedish customers had commensurately varied requirements. Meeting these diverse requirements refreshed and extended Ericsson’s stock of competencies, moved the Erieye technologies further along their respective trajectories and positioned Erieye for future opportunities. (This thesis has estimated that in course of doing so Ericsson/SAAB might have spent in the order of SEK 3000 million producing 18 Erieye systems tailored to the needs of six disparate export customers). Swedish corporatism was not confined to support for Erieye design, development and production. It also extended to support for Erieye exports. For example, the SwAF helped the Greek Air Force embed Erieye into the Greek airborne surveillance system by leasing it two SwAF Erieye systems and providing training pending delivery of the systems bought by Greece.
Swedish governance facilitated the exchange of procurement-related information between an essentially autonomous FMV and a technologically competent Ericsson Microwave Division. Hence this distinctive feature of Swedish governance reduced the time taken to develop Erieye by expediting an informed response by FMV and Ericsson to the SwAF requirement for a rapid reaction surveillance capability: Ericsson was able to establish the feasibility of adapting the PS-46A radar to meet the SwAF requirement within two years of being approached by FMV. Swedish governance reduced the time taken to develop Erieye by improving the efficiency and effectiveness with which FMV and Ericsson Microwave Division exchanged the large amounts of data required to determine how best to adapt the PS-46A radar to meet the SwAF requirement. Swedish governance permitted trusted actors in FMV and Microwave Division to exchange data in uncodified form. This enabled the two parties to determine which configuration of PS-46A radar and electronic scanned arrays represented best value for money within three years of the testing of the Erieye technology demonstrator in 1990. Thirdly, Swedish governance reduced the time taken to develop Erieye by enabling FMV-based capability entrepreneurs to take advantage of FMV’s in-house technological expertise, its knowledge of Ericsson capabilities and its autonomy. For example, by ordering long lead items in anticipation of the final decision to proceed with the SwAF procurement of Erieye, Lonroth reduced the lead time for transition from Erieye technology demonstrator to Erieye production by an estimated 18-24 months.
Swedish governance also facilitated the diffusion of Erieye overseas after completion of deliveries to the SwAF. FMV’s willingness and ability to facilitate the sale on a government- to-government basis complemented and reinforced Ericsson/SAAB’s ability to market the Erieye system to such non-Swedish customers as, for example, Thailand. In helping broker such government-to-government sales, FMV was able to draw on detailed understanding of Erieye’s capability value in SwAF service, detailed technical knowledge of the Erieye system
110
gained through involvement in developing and procuring the system and, finally, detailed understanding of Ericsson/SAAB’s technical, managerial and commercial capacity.
6.2: The influence of the Swedish defence competence bloc The informed customer function in the Swedish defence competence bloc is performed by the Minister for Defence, the Headquarters of the Swedish Armed Forces, FMV, FOI and the SwAF. During the Cold War, the efficiency and effectiveness with which this composite entity performed the informed customer function was influenced by strongly held and generally accepted agreement among the actors involved that a Soviet airborne assault constituted the primary threat to Swedish security. Such general acceptance of the nature and scale of the Soviet threat was complemented by equally general agreement that the doctrine of ground-based air defence constituted Sweden’s best option of countering that threat.
The Swedish customer actors took some 10 years to convert a requirement for a rapid reaction surveillance capability into demand for a robust but affordable ‘make’ solution. The ‘make’ solution was based on a synthesis of the proven PS-46A radar and emergent electronically scanned array technologies. Thereafter, however, Swedish capability and commercial entrepreneurs worked to ensure the Swedish defence competence block progressed the development efficiently and effectively. By 1980 (two years after being commissioned to do so), Ericsson Microwave Division had established the feasibility of this solution. Five years later, in 1985, actors in the customer element of the Swedish competence bloc had settled the parameters of a solution based on this synthesis and had agreed to proof of concept (begun in 1985 and completed in 1990 at a cost of some SEK 43 million). The Swedish customer then took a further three years to select a platform for the Erieye system and to choose the ground-based data processing option, enabling Ericsson Microwave Division to begin series production of Erieye in 1993.
Swedish capability and commercial entrepreneurs took some 21 years to take the Erieye system from feasibility study (begun in 1978) to final delivery to the SwAF (completed in 1999). This record can be attributed partly to the dense networks between, on one hand, the actors performing the customer function and those actors performing other functions of the Swedish defence competence bloc. The Swedish customer’s consistent emphasis on ‘make’ solutions to evolving capability requirements and the legitimacy of corporatist links between that customer and other elements of the Swedish competence bloc caused Swedish networks and information exchange to co-evolve. During the search for a solution to the SwAF requirement for a rapid reaction surveillance capability, pre-existing networks were such that the Swedish customer could exchange large volumes of requirements- related information in uncodified form with other actors in the competence bloc. When the selection of a solution to the SwAF requirement required the exchange of increasingly complex and detailed technical information about, for example, cost-capability trade-offs,
111
the existing networks readily adapted to enable the customer and supplier to exchange such information with minimal codification. Dense, adaptable networks within the customer element of the Swedish competence bloc, and between that customer element and the industrialist element of the competence bloc, enabled Erieye to go from proof of concept to series production within three years. Customer-centred links also facilitated post- acceptance diffusion of Erieye. One such linkage was the informed support provided by the Swedish customer to Ericsson/SAAB in their marketing of the Erieye system to non-Swedish customers. Such informed defence customer support was critical to the sale of the system to, for example, Greece and Thailand.
Swedish actors performing the innovator function were primarily located in the Swedish private sector. Specifically, Sweden’s radar-related innovators were primarily located in Ericsson Microwave Division and, after that Division’s sale to SAAB in 2006, in SAAB Electronic Defence Systems. This section analyses how Erieye innovation outcomes were influenced by links between Ericsson Microwave Division engineers and a series of other actors in the Swedish defence innovation system. These other actors included Swedish technology enablers (notably the Physics Department of Chalmers University), Swedish demand actors (notably FMV), Swedish industrialist actors (notably SAAB and CELSIUS), non- Swedish component suppliers (notably US micro-electronics suppliers) and non-Swedish customers for Erieye (notably Brazil, Greece and Pakistan).
Close collaboration between Swedish defence suppliers and Swedish universities is a distinctive feature of the Swedish defence innovation system. Such collaboration was fostered by, and contributed to, dense networks between practising engineers employed by Swedish defence suppliers and research-oriented scientists based on a common technological interest.212 Such collaboration influenced both the trajectory followed by Erieye innovation and the time taken to move along that trajectory. For example, collaboration with Chalmers scientists during the 1970s enabled Ericsson engineers to gain sufficient knowledge of electronically scanned array technology to argue successfully for its use in adapting the PS-46A radar to meet the SwAF requirement for a rapid reaction surveillance capability. A decade later, similar collaboration with Chalmers enabled Ericsson to move along the phased array technological trajectory and to develop and test an active phased array prototype relatively quickly (in less than five years) and then field an Erieye technology demonstrator relatively cheaply (FMV paid Ericsson SEK 43 million). Chalmers’ collaboration with Ericsson enabled the latter to field an active phased array prototype in time to exert a decisive influence on FMV’s selection in the mid-1980s of a solution to the SwAF requirement based on a combination of the relatively new active phased array technology and the proven PS-46A pulse Doppler radar technology.
212 Aase Jacobsson, interview, Fairbairn, 15 January 2010. 112
The dense networks that characterised the Swedish defence competence bloc enabled FMV to monitor the Chalmers/Ericsson work on phased array technology. Such technology- oriented networking between Ericsson-based innovators and FMV-based customers intensified during 1985-90 when the Erieye technology demonstrator was developed and tested in the finalisation of the Erieye design in the lead up to series production in 1993. The process of finalising the Erieye design captured the trade-offs involved in settling such issues as the range at which the system could detect small fleeting targets at what cost.
Settling these issues required the actors involved to exchange huge amounts of complex information efficiently and effectively. This was achieved by relying on the exchange of information in largely uncodified form, with information only codified to the extent necessary to capture choices and decisions. Ericsson-based innovators and FMV-based customers had built the requisite networks over successive procurement projects, a process facilitated by a common engineering education and, often, a common SwAF experience. Such dense networks between Ericsson-based innovators and FMV-based customers enabled them to finalise the Erieye design within three years of completing the testing of the Erieye technology demonstrator.
Erieye innovation outcomes were also influenced by the efficiency and effectiveness of the design-mediated networks. These networks linked Ericsson Microwave engineers to the SAAB engineers (responsible for modifying the SAAB 340 aircraft platform for the sensor system) and to the CELSIUS engineers responsible for integrating it into the STRIL/STRIC command and control system. Design-mediated communication across those networks was orchestrated by FMV, who assumed prime contractor responsibilities in managing the procurement of the sensor systems, the procurement, modification and fit-out of the SAAB 340 platforms and integration of platform/radar systems into Sweden’s ground-based air defence system. The work of designing and building the Erieye radar, of designing and carrying out the modification/fit-out of the SAAB 340 aircraft and of integrating the radar and its transport platform into the STRIC/STRIL system took some nine years, beginning after the successful testing of sensor technology demonstrator in 1990 and continuing in parallel with series production of the system during the 1993-99 period. The interrelated competencies established in the course of this learning-by-doing enabled the requirements of non-Swedish companies to be met much more quickly.
FMV procured six Erieye systems for the SwAF at an estimated total project cost of SEK 1,243 million. Compared to other airborne early warning aircraft, this represented remarkable value for money. Ericsson’s ability to achieve such value for money was partly due to the company’s ability to produce the electronic scanned array at an acceptable cost. This was in turn largely attributable to Ericsson’s ability to access US microelectronic technology,213 initially under the provisions of the Stockholm Agreement concluded by the
213 Lars Karlen and Roland Karlsson, interview, Molndahl, 20 November 2009. 113
US and Swedish governments, and subsequently under the revised understandings between the two governments concerning Swedish protection of US technology concluded in May 1982.
This sub-section analyses how action by Swedish capability entrepreneurs (notably FMV’s Carl-Gilbert Lonroth) and commercial entrepreneurs (notably engineers in Ericsson Microwave Division) affected the time taken to develop ERIEYE, the cost incurred in doing so and its pattern of development following its acceptance into service by the SwAF. The time taken to develop ERIEYE was influenced by Lonroth’s judgement that, at a time when the Swedish defence customer was preoccupied with developing the Gripen fghter, it had little appetite for the cost, schedule and technical risk inherent in de novo development of a rapid reaction surveillance radar. In initiating the search for a solution to this requirement, Lonroth was well aware of the capabilities of the pulse Doppler radar technologies used in the proven PS-46A radar and was predisposed in favour of a path-dependent development of that technology. Lonroth was also aware of the work on electronically scanned arrays by Chalmers and Ericsson.
In commissioning feasibility studies by Ericsson in 1978-80, Lonroth was taking an entrepreneurial initiative based on an informed judgement about the applicability of these two technologies to the SwAF requirement. Lonroth took the initiative with a view to codifying otherwise largely tacit knowledge he exchanged with other Swedish defence actors – including Westerlund and the Ericsson engineers – with a stake in solving the SwAF requirement. In commissioning the studies in 1978, Lonroth did not influence how long Erieye development took, what it cost or what happened after it entered SwAF service. Rather, by setting in train a sequence of selection and procurement action, Lonroth’s initiative largely defined when these derivative actions occurred. Because acceptance by the SwAF was a precondition for export, Lonroth’s initiative also defined when post acceptance development of Erieye began (but not, of course, the precise timing or incidence of export sales).
Ericsson Microwave Division acted entrepreneurially in contributing some SEK 43 million to the cost of developing a rapid reaction surveillance radar with a range far in excess of the 60 km range that the SwAF sought and was prepared to fund. In taking this entrepreneurial initiative Ericsson engineers judged that the potential non-Swedish market for a relatively low-cost, high-capability broad area surveillance system warranted Ericsson meeting the extra cost of more fully exploiting the system’s potential as indicated by the 1978-80 feasibility studies. Such entrepreneurialism was entirely consistent with the Ericsson company’s broader commercial behaviour (particularly in the intensely competitive international telecommunications market). It was also compatible with the ethos fostered by Marcus Wallenberg who was Chairman of the Ericsson Board and who had a controlling interest in Ericsson through Investor AB. Ericsson Microwave Division’s entrepreneurialism
114
was a prerequisite for the diffusion of Erieye technology that occurred after the Division had completed Erieye deliveries to the SwAF in 1999.
According to Lonroth, Ericsson spent about the same as FMV – some SEK 43 million – in building a demonstrator with an instrumented range of well over 300 km. This not only satisfied the SwAF operational requirement but created opportunities to meet the needs of non-Swedish customers with much more demanding requirements. As already indicated, Ericsson’s assessment of the commercial potential of this entrepreneurial initiative was reinforced by the Swedish policy of giving Ericsson ownership of Erieye IP and by the prospect of FMV/SwAF support for export of the system. Such exports were, however, restricted to customers that complied with, firstly, Swedish arms export control policies and with, secondly, US International Traffic in Arms Regulations (because Erieye incorporated US micro-electronics). Within these constraints, Ericsson’s entrepreneurial initiative cleared the way for the company to begin exporting Erieye to Brazil in 1998, a year before it had completed deliveries to the SwAF.
Research for this thesis found nothing to suggest that action taken by Marcus Wallenberg, Investor AB, or the Stockholms Enskilda Banken (and, after its merger in 1972, the Skandinaviska Enskilda Banken) directly and immediately influenced the time taken to develop Erieye, the cost incurred in doing so or the pattern of Erieye development after its acceptance by the SwAF. Olsson, for example, makes no mention of Erieye in either his biography of Wallenberg or his history of the Wallenberg-owned and -controlled Stockholms Enskilda Banken and its successor, the Skandinaviska Enskilda Banken.214 As part of the Wallenberg Group of companies, Ericsson and, by extension, Ericsson Microwave Division had access to the kind of patient capital that innovators need to develop, for example, electronically scanned arrays for both airborne and ground-based radar applications. It can also be argued that, as a result of Wallenberg’s personal interest in advanced technology and of his interventionist management style, the Ericsson Board would have appointed managers of the Ericsson Microwave Division able and willing to pursue such long-term investments and to bring them to a commercially successful conclusion.
As innovators, Ericsson Microwave Division engineers took over two decades (from 1968 to 1990) to take electronically scanned array technology from the experimental stage through to production of a prototype active phased array. As commercial entrepreneurs, Ericsson Microwave Division engineers only stood to make a commercial return on this protracted niche development in the early 1990s, when FMV placed a contract with Ericsson for the series production of electronically scanned arrays for, initially, ground-based battlefield surveillance radars and later, Erieye.
214 See Ulf Olsson, At the Centre of Development: Skandinaviska Enskilda Banken and its predecessors 1856- 1996, Skandinaviska Enskilda Banken, Stockholm, 1997; and Ulf Olsson, Furthering a Fortune: Marcus Wallenberg Swedish Banker and Industrialist 1899-1982, Ekerlids Forlag, Stockholm, 2001. 115
Funding such protracted development would have required the support of the Ericsson Board. That Board, which was chaired by Marcus Wallenberg, was the focal point of a dense commercial and managerial network connecting the engineers in Ericsson Microwave Division to the venture capitalists in Investor AB. Therefore, by indirectly funding the niche development of a critical technological input to Erieye, Marcus Wallenberg and Investor AB established the technological and managerial preconditions for Erieye development and production. By being able to judge the commercial potential of electronically scanned array technology, the Ericsson Board led by Wallenberg provided the long-term support that Ericsson Microwave Division needed in developing and producing a solution to the SwAF rapid reaction surveillance requirement within 10 years of confirming that a combination of proven pulse Doppler radar technology and emergent electronically scanned array technology was feasible.
Up until the Division’s sale to SAAB in 2006, Ericsson Microwave Division engineers performed the Erieye industrialist function as well as the Erieye innovator and entrepreneur functions. In doing so they drew on radar-related competencies established and maintained by Ericsson Microwave Division through design, development and production of successive generations of radar for the SwAF and other customers. Hence, for example, the skills required to design Erieye were a path-dependent extension of those acquired by Ericsson engineers in designing and developing a pulse Doppler radar for the Viggen aircraft. This protracted and expensive learning process began in 1963 and culminated 15 years later in the production of a prototype PS-46A radar in 1978. That said, the skills so developed enabled Ericsson engineers to finalise the Erieye design much more quickly – producing an Erieye technology demonstrator within five years of the decision to do so in 1985 and finalising the design for series production in 1993, three years after testing the prototype. Similarly, despite Ericsson Microwave Division engineers having to master new electronically scanned array technology, they succeeded in producing, on a fixed price basis, the six Erieye procured for the SwAF in six years. This expeditious production performance can be partly attributed to production skills developed during production of the PS-46A radar (1978-87) and, from 1990, production of the PS-05A radar for the Gripen aircraft. The above Erieye- related industrialist competencies enabled Ericsson Microwave to identify and meet overseas demand for various configurations of Erieye equally expeditiously. This influenced the pattern of Erieye development post-SwAF acceptance decisively: during the period 1998-2013, Ericsson Microwave Division and its successor sold some 18 systems to six countries with widely varying requirements.
6.3: The influence of Swedish military doctrine This section analyses how Sweden’s doctrine of ground-based air defence affected the time taken to develop ERIEYE, the cost incurred in doing so and the pattern of Erieye development after its delivery to the SwAF. For the first 20 years of the Cold War, the SwAF used ground-based microwave radars configured for search, targeting and height finding to
116
execute the surveillance element of its ground-based air defence doctrine. The SwAF requirement for a rapid reaction surveillance system only emerged in the late 1960s, when Soviet tactical innovations reduced the utility of the SwAF’s ground-based microwave radars in providing timely detection of, and response to, attacking Soviet aircraft.
The SwAF’s strong commitment to its distinctive ground-based air defence doctrine confined demand for a solution to the SwAF rapid reaction surveillance requirement to the existing microwave radar technological paradigm and the prevailing manned fighter intercept operational paradigm. Such a solution took nearly a decade to emerge. It was not until 1978 that Lonroth commissioned Ericsson to investigate the feasibility of combining a pulse Doppler microwave radar with an electronically scanned array and mounting the whole package on board an aircraft able to transport it at the requisite altitude and for the requisite time. In acting as a capability entrepreneur within the constraints of SwAF air defence doctrine, Lonroth took advantage of a combination of technological opportunity and economic opportunity.
The technological opportunity was created by Ericsson and Chalmers’ successful development of electronically scanned array technology. This development took over 20 years, beginning with an experimental electronically scanned array in 1968 and ending with a prototype active phased array in 1990. The technological opportunity was enhanced by investments in the capacity of the Swedish command and control system. These investments included enhanced ground-to-air data links. These helped the SwAF fighter pilots use the much larger volumes of data generated by the electronically scanned array technology when fused with the PS-46A platform. Enhanced links also enabled the intercepter to achieve tactical surprise by obviating the need for the intercepting aircraft to illuminate the target, thereby alerting the target to the interceptor’s presence.
The economic opportunity emerged in the 1980s, after the Viggen program had absorbed the costs of PS-46A development, which could be treated as sunk costs for Erieye program purposes. FMV and the SwAF had a strong incentive to exploit this opportunity because, at the time, Swedish planners had little appetite for ab initio development of a new surveillance radar. Sweden’s commitment to the Gripen program in 1980 and the easing of strategic tensions after 1985 (as a result of Gorbachev’s Glasnost program and the rapprochement between the US and the USSR) both reduced resources for, and reduced the urgency of, meeting the requirement for a rapid reaction surveillance capability. Nevertheless, in the lead up to the dissolution of the USSR in 1991, Swedish planners perceived the need to hedge against USSR/Russian resurgence by providing early warning of any airborne assault.
As already indicated, development and production of six Erieye cost about SEK 1,243 million (excluding Ericsson subsidy). In unit cost terms, this is approximately 25% of the unit cost of
117
a Grumman E2D Hawkeye.215 As such, the Erieye solution represented outstanding value for money in meeting the SwAF requirement for a rapid reaction surveillance requirement within the framework of the prevailing SwAF ground-based air defence doctrine. At a unit cost of some SEK 207 million, Erieye was a relatively economical solution. This was made possible by, firstly, the sunk cost of PS-46A development already mentioned. The Erieye program also benefitted from Ericsson Microwave Division’s investment in electronic scanned array technology which had been prompted by both ground-based radar requirements and Gripen’s multirole radar requirement.
Erieye was effective in the sense that it met the SwAF operational requirement with an aircraft that, because it weighed about 25% of the Hawkeye and had a shorter wingspan, could operate from SwAF road-based runways. Erieye’s effectiveness was enhanced by using ground-based operators and data links that integrated seamlessly into SwAF’s ground- based air defence system. Finally, by drawing on pre-established competencies developed by both FMV and Ericsson Microwave Division, Erieye complied with Sweden’s preference for ‘make’ solutions based on demonstrably ‘neutral’ technology. Erieye was also an efficient solution in sense that Ericsson Microwave Division delivered it to the SwAF on time and to cost. This was due to the common understanding about Erieye cost, schedule and technical risk established between that Division and FMV through Erieye feasibility studies and the technology demonstrator program. The efficiency of the Erieye solution was enhanced by the ability of Ericsson Microwave Division to adjust its stock of Erieye-related competencies in the course of these activities.
Sweden’s ground-based air defence doctrine and the SwAF’s requirement for a relatively cheap but effective rapid reaction surveillance system to give effect to that doctrine constituted a distinctive response to Sweden’s unique Cold War circumstances. Nevertheless, the Erieye solution to the SwAF requirement was attractive to other countries which, like Thailand and Pakistan, had comparable air defence doctrines and, like Brazil, comparable broad area surveillance requirements. Erieye’s appeal to these countries was enhanced by the ability of Ericsson/SAAB to adapt the system to varying customer requirements efficiently and effectively. The upshot was that within 14 years of completing deliveries to the SwAF in 1999, Ericsson/SAAB succeeded in selling variants of the Erieye system to at least six countries.
215 Estimate based on USN payment of $US617 million for five full rate production versions of the E2D Hawkeye in 2013, see http://www.navyrecognition.com/index.php?option=com_content&task=view&id=1185, accessed 7 February 2014.
118
6.4: The influence of the Swedish technology base Sweden’s distinctively configured technology base was a product of several convergent influences. These included the SwAF’s ground-based air defence doctrine, Swedish demand for neutral technology, the high value accorded by Swedish defence actors to indigenous ‘make’ solutions and an informed customer’s insistence that any such solutions provide a credible defence against a Soviet airborne assault. Sustained Swedish demand for a militarily competitive ground-based air defence system fostered a technology base populated by densely integrated technological systems. The latter included not only radar technological systems (the main focus of the present discussion) but also platform technological systems; command, control and communication technological systems; munitions technological systems; and infrastructure technological systems. The Swedish air defence technology base can therefore be viewed as a system of technological systems that co-evolved along a path shaped by both endogenous and exogenous influences. Against this background, this section analyses how the evolution of Sweden’s radar technological system affected the time taken to develop Erieye, the cost incurred in doing so and the pattern of Erieye development after its delivery to the SwAF.
During the Cold War, the evolution of Soviet airborne assault capabilities underpinned sustained demand by Swedish defence actors for a militarily competitive air defence system. This demand, shaped by the SwAF’s air defence doctrine and focused by the high value accorded technologically neutral, indigenous ‘make’ solutions, prompted L.M. Ericsson (with the support of the Wallenberg financial institutions) to invest in sustained development of a Swedish radar technological system. The co-evolutionary nature of the Swedish technology base meant that, once established, the Swedish radar technology system developed along a trajectory defined by a combination of demand-pull and technology-push influences. On the demand-pull side, Ericsson innovators and entrepreneurs responded to demand as formulated by an informed customer (which encouraged Ericsson to push development of the Swedish radar technological system along the pulse Doppler radar technology trajectory). On the technology push side, Ericsson innovators and entrepreneurs anticipated demand by FMV for enhanced radar beam forming to enable the SwAF to undertake adaptive search for, identification of, tracking of and targeting of incoming Soviet aircraft. Anticipating such demand prompted Ericsson innovators and entrepreneurs to push development of the Swedish radar technological system along the electronically scanned array technology trajectory.
By the time an FMV-based capability entrepreneur took action to address the SwAF requirement for a rapid reaction surveillance capability, Ericsson innovators and entrepreneurs had pushed the Swedish radar technology system sufficiently far for an acceptable solution to emerge to meet the SwAF requirement. In these circumstances, the initial task for Ericsson innovators and entrepreneurs was to establish the feasibility of synthesising the available technological precursors to meet the SwAF requirement within
119
parameters of cost, schedule and technical risk that both the potential customer and developer deemed acceptable. In doing so, Ericsson Microwave Division engineers prepared the Swedish radar technological system for a speciation event causing that system to develop along a new technological trajectory that the actors involved could envisage in broad outline though not in every specific detail. Once Ericsson innovators and entrepreneurs had established the feasibility of meeting the SwAF requirement by a synthesis of known or knowable technological precursors, their next task was to design an artefact that demonstrated the capability of the synthesis and that pointed the way for further refinement of the artefact in light of knowledge gained through trialling and testing it. In doing so, Ericsson Microwave Division engineers and FMV-based capability entrepreneurs established a technology niche within which the above speciation took place, laying the foundations of the new technological trajectory to be followed by the Swedish radar technological system.
Subject to the customer’s decision to procure a version of the artefact, Ericsson innovators, entrepreneurs and industrialists then further refined and embellished the artefact to obtain better value for money. In doing so, they consolidated the new technological trajectory to be followed by the Swedish radar technological system until capability entrepreneurs and commercial entrepreneurs identified new technological and economic opportunities that, ex ante, pointed to further movement along the existing trajectory (for example, by greater range, sensitivity or precision) or to a shift to a new trajectory (involving, for example, a new function or a new technology).
The ability of Swedish capability entrepreneurs and commercial entrepreneurs to utilise already available precursor technologies reduced the time taken to develop a solution to the SwAF requirement. Specifically, the availability of pulse Doppler radar technology and of electronically scanned array technology in the radar technological system element of Sweden’s technology base enabled Ericsson Microwave Division engineers to field an Erieye technology demonstrator within five years of the decision (in 1985) to develop one. The performance of that demonstrator was sufficiently close to meet the SwAF requirement for a rapid reaction surveillance capability to enable series production of the Erieye system to begin in 1993, a mere three years after testing of the technology demonstrator had been completed. However, the Ericsson Microwave Division engineers could only achieve the development schedule by harvesting extensive and protracted prior investments in learning- by-doing. One tranche of such learning-by-doing centred on developing the pulse Doppler radar technology used in PS-46A radar for the Viggen aircraft. The other tranche of learning- by-doing centred on developing the electronically scanned array technology without which the PS-46A pulse Doppler radar could not be adapted to meet the SwAF requirement for a rapid reaction surveillance capability.
120
Ericsson Microwave Division took some 12 years to master pulse Doppler radar technology sufficiently to meet the SwAF’s requirement for a look down/shoot down capability. This learning-by-doing comprised two years of initial experimentation (1963-65), a further two years of more focused experimentation (1965-67) followed by a further eight years of developing the prototype PS-46A radar. Thereafter, Ericsson Microwave Division took an additional seven years to adapt the PS-46A radar to the SwAF’s rapid reaction surveillance requirement. Adapting the radar took two years of feasibility studies (1978-80) and five years of technology demonstration (1985-90).
The second, equally important tranche of learning-by-doing related to the investment by Ericsson Microwave Division and Chalmers University in developing electronic scanned array technology. Despite Chalmers’ support, Ericsson Microwave Division took 22 years to develop the electronic scanned array technology to the point at which it could be used for airborne early warning and control purposes. This development included three years of initial experimentation (1968-71), followed by nine years of technology demonstration (1971-80) and 10 years of prototyping (1980-90).
The cost of developing artefacts like the Erieye system lies not in the fabrication of the artefact but in developing, marshalling and applying the knowledge required to design and build the artefact. The direct cost to FMV of designing and building the Erieye technology demonstrator was SEK 43 million. For this relatively modest sum, Ericsson Microwave Division was able and willing to build an Erieye technology demonstrator able to identify small fleeting targets like fighter aircraft and cruise missiles at ranges of 300-400 km (some five times that specified by the SwAF). The requisite radar-related knowledge was only available in Ericsson Microwave Division because of the cumulative learning gained through that Division’s involvement in previous radar-related programs. Two points are relevant here. Firstly, while the cost of such cumulative learning had largely been absorbed by previous programs undertaken by Ericsson Microwave Division, Swedish corporatist norms meant that the stock of knowledge so accumulated remained available at little or no cost to Ericsson Microwave Division personnel assigned to the Erieye development task. Secondly, only a modest increment of knowledge was required to go from the level Ericsson Microwave Division innovators had established by developing pulse Doppler radar and electronically scanned array technologies to that they needed in order to combine these technologies to meet the SwAF requirement.
The total cost of the Erieye technology demonstrator was SEK 80-90 million, comprising SEK 43 million paid by FMV plus a similar amount contributed by Ericsson Microwave Division in return for full Erieye IP rights. This sum, equivalent to 6-7% of the SEK 1,243 million paid by FMV for development and production of Erieye for the SwAF, is an indication of the marginal cost of moving the Swedish radar technological system sufficiently far down pulse Doppler radar and electronically scanned array technological trajectories to meet the SwAF
121
requirement for a rapid reaction surveillance capability. Expenditure by FMV and Ericsson of the above SEK 80-90 million to develop a technology demonstrator positioned the Swedish radar technological system for entry into the international market for microwave radar- based broad area surveillance systems in the late 1990s. Such expenditure, while a necessary condition for such entry, was by no means a sufficient condition. A distinctive feature of the international market for military technological innovation (particularly by smaller countries like Sweden) is that purchase of an innovation by the military of the country in which the developer is based is essentially a prerequisite for sales abroad. The military utility of the fusion of pulse Doppler radar technology and electronically scanned array technology engineered by Ericsson Microwave Division innovators and entrepreneurs was confirmed by sales of Erieye to the SwAF in 1993-99. The capability value generated by Erieye once the system was embedded into an appropriately configured portfolio of air defence assets was demonstrated by the SwAF’s incorporation of Erieye into Sweden’s ground-based air defence system.
In marketing Erieye to non-Swedish customers, Ericsson entrepreneurs were able to draw on knowledge of the military utility of Erieye in detecting small fleeting targets in an electronically cluttered environment. The ability of Ericsson (and later SAAB) to export Erieye profitably was enhanced by FMV’s payment of SEK 43 million toward the initial cost of the system’s development and by FMV’s payment for Ericsson’s learning-by-doing in the course of producing six units for the SwAF. The ability of Ericsson (and later SAAB) to export Erieye in the face of stiff competition from the US and other suppliers was enhanced by the companies’ ability to leverage Sweden’s comprehensive, densely integrated radar technological system in adapting Erieye to particular customer requirements. The upshot was Ericsson/SAAB’s sale of some 18 systems (three times the number sold to the SwAF) to six countries over the period 1998-2011.
6.5: The influence of Swedish demand The demand element of the Swedish defence sectoral innovation systems comprehends the process by which Swedish defence actors devise a novel solution to a requirement for a new military capability through a process of search, selection and procurement action. This section analyses, firstly, how the process by which Swedish defence actors searched for a solution to the SwAF’s requirement for a rapid reaction surveillance capability affected the time taken to develop ERIEYE, the cost incurred in doing so and the pattern of Erieye development after its delivery to the SwAF. This is followed by an analysis of how the process by which those actors selected a solution to the SwAF’s requirement for a rapid reaction surveillance capability affected the time taken to develop Erieye, the cost incurred in doing so and the pattern of Erieye development after its delivery to the SwAF. The section concludes with an analysis of how the process by which the actors procured a solution to the SwAF’s requirement for a rapid reaction surveillance capability affected the
122
time taken to develop Erieye, the cost incurred in doing so and the pattern of Erieye development after its delivery to the SwAF.
The search for a solution to the SwAF’s rapid reaction surveillance requirement was initiated by Carl-Gilbert Lonroth, a capability entrepreneur based in FMV. There was nothing random about the search process. The parameters of FMV’s search for a solution to the SwAF’s rapid reaction surveillance requirement were set by the SwAF’s ground-based air defence doctrine and the associated commitment to use of manned fighter aircraft controlled by ground-based command centres to intercept incoming Soviet aircraft. Within these parameters, the search was further conditioned by the prevailing radar-based technological paradigm and, within that paradigm, the success of Sweden’s innovators, entrepreneurs and industrialists in leveraging Sweden’s microwave radar-based technological system in meeting the SwAF’s continually evolving requirement for militarily competitive sensor systems. It is therefore reasonable to argue that the search initiated by Lonroth represented a path-dependent extension of the established development trajectory of Sweden’s airborne microwave radar system. Lonroth’s search for a solution to the SwAF’s rapid reaction surveillance requirement comprised two closely linked but sequential phases. The first search phase consisted of the two-year program of feasibility studies conducted by Ericsson Microwave Division over the 1978-80 period. The second phase consisted of that Division’s five-year program to develop a technology demonstrator over the 1985-90 period. Taken together, this two-phase search for a solution to the SwAF requirement account for one-third of the 21 years (1978-99) Ericsson Microwave Division and FMV took to field an ERIEYE-based solution to the SwAF rapid reaction surveillance requirement.
The feasibility studies drew on the large body of pre-existing knowledge accumulated by Ericsson Microwave Division about pulse Doppler radar technology and about electronically scanned array technology. Hence it is reasonable to argue that the feasibility studies entailed desktop work, the cost of which was relatively trivial, and that it can be subsumed in the cost of the Erieye technology demonstrator. FMV paid Ericsson SEK 43 million to build the Erieye technology demonstrator, the second phase of the search process. As already indicated, this payment substantially understates the full cost of the search phase which, when Ericsson’s contribution is included, is more likely to be in the order of SEK 80-90 million. For case study purposes, however, only the FMV payment of SEK 43 million is attributed to the cost of the search phase.
The search process initiated by Lonroth was geared primarily, but not entirely, to SwAF requirements. Commercial entrepreneurs based in Ericsson Microwave Division and working on the SwAF feasibility studies had already identified comparable requirements among potential non-Swedish customers. Hence the information generated by the SwAF-oriented search process also informed Ericsson Microwave Division’s assessment of the non-Swedish market for any successful solution to the SwAF requirement. To this extent, the search
123
process prepared the way for diffusion of the Erieye innovation after its acceptance into SwAF service. The actual timing of that diffusion was subject to potential buyers being able to see the system in operation by the SwAF. The latter did not begin until 1993, well after completion of the search process. Therefore, the search process did not contribute directly to the post SwAF diffusion of Erieye.
The Erieye selection process did not start until the Erieye technology demonstrator had been successfully tested (which was completed by 1990). Thereafter, Swedish corporatist norms and Ericsson’s proven record in supplying the SwAF with militarily competitive radars meant that the Erieye selection process did not add significantly to either the Erieye development schedule or to Erieye costs. As already indicated, post SwAF development of Erieye for overseas customers was dependent on successful introduction of the system into SwAF service (which did not begin until 1993). Hence it is reasonable to argue that the timing and value of Erieye exports were not directly affected either by the process by which Erieye was selected to meet the SwAF requirement for a rapid reaction surveillance capability or by the process by which Ericsson was selected to produce that system.
Procurement of the Erieye system for the SwAF was subject to successful testing and trialling of the technology demonstrator, completed in 1990. Thereafter, the process by which the customer procured Erieye for the SwAF can be divided into preparatory and production phases. The preparatory phase of the Erieye procurement process began in 1990. This phase, which lasted some three years, encompassed finalisation of the design, obtaining budget approval for the procurement of six systems, ordering long lead items and negotiating a contract for production of the system. The production phase of the Erieye procurement process began in 1993. Ericsson Microwave Division took on average a year to produce each system and completed delivery of the six systems procured for the SwAF by 1999. Taken together, the preparatory and production phases of the Erieye procurement process took a total of nine years (1990-99) – see Chapter 7, Table 7.3. As already indicated, FMV concluded a fixed price contract with Ericsson for design and production of six systems, based on the Erieye technology demonstrator, for a total cost of SEK 1,200 million.
The SwAF began accepting Erieye into service in 1993. Successful conclusion of the production phase of the Erieye procurement process prepared the way for Ericsson’s ability to export Erieye depended on the company’s ability to demonstrate both the military utility of the system and its capability value to potential overseas buyers. As the first buyer and operator of Erieye, the SwAF was critical to both endeavours: by demonstrating the system at home and abroad, the SwAF was able to corroborate Ericsson’s claims about the system’s military utility. By demonstrating how it had embedded Erieye into Sweden’s ground-based air defence system, the SwAF was able to illustrate Ericsson’s claims regarding the system’s potential capability value as an element of the potential customer’s overall air defence system. The Erieye procurement process fostered a combination of Erieye value for money,
124
Ericsson producer expertise and SwAF user experience that underpinned the diffusion of the Erieye system to six non-Swedish customers in the period 1998-2011.
6.6 Conclusion Chapter 3 of the thesis described the features of a generic defence sectoral innovation system and identified several subsidiary research questions flowing from that description. The following paragraphs address these subsidiary questions in terms of what the development of Erieye indicates about the performance of the Swedish innovation system.
Long-standing Swedish institutions (armed neutrality, corporatism and agency autonomy) preconditioned the Swedish defence sectoral innovation system for development of a novel solution to the SwAF requirement for a rapid reaction surveillance capability efficiently and effectively. These institutions fostered a Swedish defence competence bloc populated by a suite of actors, each of which maintained specialist competencies which, when linked together, were proximate to those required to produce a novel solution to the SwAF requirement for a rapid reaction surveillance capability. This proximity between competencies available as a result of past choices and competencies required to meet future choices was largely due to the dynamic stability of Sweden’s ground-based air defence doctrine. Such dynamic stability favoured path-dependent requirements for military capability during the Cold War, thereby encouraging actors to accumulate deep knowledge of the associated technological systems. Dense links among technologically specialised actors combined with an informed customer that consistently favoured ‘make’ solutions to military requirements created a diverse, adaptive Swedish technology base. The technological plurality of Sweden’s technology base, combined with Swedish actors’ ability to access overseas (especially US) innovations enabled Swedish innovators, entrepreneurs and industrialists to respond efficiently and effectively to the Swedish customer’s demand for an effective but affordable solution the SwAF requirement for a rapid reaction surveillance capability. The competencies so established enabled those same innovators, entrepreneurs and industrialists to meet comparable non-Swedish requirements in the face of stiff international competition for the business.
125
Chapter 7: The Australian Defence Sectoral Innovation System This chapter is the first of the three chapters in the thesis that examine the structure, conduct and performance of the Australian defence sectoral innovation system as part of a wider comparison of the performance of the Swedish and Australian defence sectoral innovation systems. That wider comparison is in turn part of the investigation undertaken in this thesis of why the cost, schedule and diffusion of comparable innovations in similar countries can vary widely This chapter describes the structure of the Australian defence sectoral innovation system in terms of the five building blocks described in Chapter 3. Accordingly, this chapter begins with an outline of the norms that conditioned innovation choices by Australian actors in distinctive ways. This is followed by a discussion of how the actors that populate the Australian defence sectoral innovation system perform the defence competence bloc functions. Then discussions of, respectively, Australian military doctrine and the Australian technology base follow. The chapter concludes with a discussion of how Australian defence actors execute demand for novel solutions to Australian military requirements.
7.1: Australian Institutions During the Cold War, the structure and operation of Australia’s defence sectoral innovation system was influenced by three deep-seated institutions. The first was Australia’s historical penchant for alliance-based collective security. The second more recent institution was the emphasis on self-reliant defence of Australia. The third institution, which can be traced back to the Federation of Australia in 1901, was open and effective competition for defence business.
In the Australian context, the logic of alliance-based collective security reflected Australian experience of World War Two and the Australian response to early post World War Two imperatives. During the Cold War, alliance-based collective security was founded on the ANZUS Treaty. It was initially manifest in Australia’s forward defence strategy but a more muted version underpinned Australia’s subsequent quest for more independent military capability and its focus on regional security arrangements. The impact of Australian alliance- based collective security thinking on Australian military technological innovation is discussed in the following paragraphs.
Pre-World War Two, Australia saw itself as an Anglo-Saxon outpost in the Asia Pacific and looked to Britain and the Royal Navy for security. During World War Two, Australia responded to the Japanese invasion of South East Asia and the South West Pacific by mobilising for immediate defence of the Australian continent and by collaborating closely with the US in rolling the Japanese back. In 1946 this experience led the Australian Chiefs of Staff to argue that: 126
1 Australia, being an isolated continent with a small population and limited resources, is unable to defend herself unaided against a major power.
2 It follows that a policy of isolation can only lead to disaster, and that her security must be based upon co-operation with other nations.
3 It further follows that –
a. Her preparations for war must be such that that her forces can cooperate with those of other nations, and
b. Her dependence on outside assistance, compels her to accept that the strategic employment of her forces will be governed by considerations wider than those of a purely regional nature.216
The above logic was reinforced when the Chinese Communist Party won the Chinese civil war in 1949. It was further reinforced when the Korean War between the United Nations and communist forces broke out in 1950 and when Commonwealth forces began engaging communist insurgents in the protracted Malayan Emergency, also in 1950. Australia’s pervading sense of communist threat was reinforced in 1954 by the Viet Minh defeat of the French in the first Indochina war and by East-West tensions in Western Europe.
In the late 1940s, the US took the lead in establishing coalitions of like-minded nations (including Japan) to counter what was widely perceived as a global communist onslaught. In return for signing a peace treaty with Japan (and thereby facilitating its inclusion in the US global anti-communist coalition), Australia and New Zealand sought a US commitment to support them in case a re-armed Japan threatened them again. The US acquiesced and in September 1951 concluded the ANZUS Treaty with Australia and New Zealand. Article IV of the treaty provides that:
Each Party recognizes that an armed attack in the Pacific Area on any of the Parties would be dangerous to its own peace and safety and declares that it would act to meet the common danger in accordance with its constitutional processes. 217
Australian government perceptions of the utility of the ANZUS Treaty evolved rapidly as initial preoccupation with Japan was displaced by the perceived threat to Australian security from Asian communism. The generally worded provisions of Article IV became the foundation for Australia’s evolving norm of collective security. Australian governments recognised, however, that the nature and scale of American support to Australia in a
216 Chiefs of Staff Committee: Appreciation of the Strategical Position of Australia, 1946 in Stefan Freuling, History of Australian Strategic Policy since 1945, Defence Publishing Service, Commonwealth of Australia, Canberra, 2009, p. 63. 217 Gary Brown and Laura Rayner: Upside, Downside: ANZUS after Fifty Years, Parliament of Australia, Canberra, August 2001, p. 4. 127
contingency would depend on how American policymakers calculated American interests at the time. To improve the likelihood of such support, Australia encouraged US engagement in the Asia-Pacific region, contributed troops to the Korean and later the Vietnam wars, and hosted US strategic communications and intelligence-gathering facilities.218
Australian perceptions of communist threat in East and South East Asia receded in the 1960s. Australian defence policymakers accorded higher priority to developing a capability for more independent military action in defence of Australian interests. This led Australian governments to place more store on Article II of the Treaty, which provided that:
In order more effectively to achieve the objectives of this Treaty, the Parties separately and jointly by means of continuous and effective self help and mutual aid will maintain and develop their individual and collective capacity to resist armed attack219.
During the 1970s, Australia sought to establish and maintain a capability for independent action at acceptable cost by accessing advanced US technology and intelligence. By the mid- 1980s Australian defence planners recognised that “only the United States could provide Australia with the intelligence, defence technology and professional military expertise which would enable Australia to independently handle regional threats.”220
Article II of the ANZUS Treaty legitimised the conclusion of hundreds of derivative government-to-government agreements intended to facilitate the transfer of more specific technology and materiel from the US to Australia. Under the agreements, the US government essentially accords Australia NATO status for the purposes of transferring advanced US technology, including for licensing US companies to export advanced – and sensitive – US military equipment and technology to Australia for use by the Australian Defence Force. Such access influenced Australia in favour of ‘buy’ solutions to its requirements for military capability during the Cold War period. During the early decades of the Cold War, these choices were also influenced by Australia’s strategy of forward defence.
In the 1950s-1960s Australian governments gave practical effect to the norm of collective security, by adopting a forward defence strategy. This was intended to provide Australia with defence in depth by preventing a potential (communist) enemy establishing itself in South East Asia from where it could threaten Australia. Commitment to forward defence entailed trading off some sovereign control of Australian armed forces for security. As the Defence Committee observed in 1968:
we have deliberately ... tied Australia to the strategy of others. We have had such a tradition, first to fit comfortably into British strategy and more recently in that of the US. In this latter case we have placed our trust in ANZUS, we played a major part in
218 ibid, p. 5. 219 Ibid, appendix 1. 220 Des Ball, The strategic essence, Australian Journal of International Affairs, Vol 55 (2), p. 236. 128
establishing (the South East Asia Treaty Organisation), of which the USA is the dominant partner. Like all small countries we can best ensure our security by participating in regional security arrangements; as a result we find ourselves involved in situations not of our choosing and in the formation of which we have negligible if any influence.221
Australia gave practical effect to its forward defence strategy by committing forces to coalition operations in South East Asia with the US, the UK and other like-minded countries. In order to facilitate Australian participation in such operations Australian defence policy stipulated that:
Equipment used by the Australian forces should be standard or compatible as far as possible with that used by United States forces.222
In the context of forward defence, Australian governments’ willingness to compromise sovereign control of Australian armed forces and Australian defence policy emphasis on interoperability with friends and allies led Australian defence planners to accord commensurately low value to ‘make’ solutions to Australian materiel requirements. Australian defence policy expressly recognised that the limited size of the Australian forces made local development and manufacture of major items of military equipment uneconomical. The policy recognised that local defence research and development was necessarily limited and, hence, should complement rather than duplicate that of Australia’s allies. While local research and development could usefully address those few problems particular to the operation of the Australian forces, local ‘make’ solutions were not to be pursued at the expense of providing the modern weapons necessary for effective armed forces.223
In the early 1960s, Indonesian ‘confrontation’ of Malaysia shifted Australian policy attention away from forward defence to the security of Australia’s immediate neighbourhood. Australian policymakers recognised that, in the event of hostilities with Indonesia, Australia could not count on the US providing, under ANZUS auspices, US armed forces in support of Australia.224 Australian recognition that the degree of US involvement would depend on the US calculation of its interests prompted the Australian government to invest in the capacity for more independent action. Such investments included, for example, ordering F111 bombers from the US under the auspices of Article II of the ANZUS Treaty. But Australia’s quest for more sovereign military capability did not entail according any higher value to ‘make’ solutions to Australian military capability requirements than hitherto.
By mid-1968 Indonesian ‘confrontation’ of Malaysia had ceased, the Indonesian Communist Party had been purged and the Association of South East Asian Nations (ASEAN) had been
221 Defence Committee’s Strategic Basis of Australian Defence Policy, August 1968, in Freuling, p. 376. 222 Defence Committee’s Strategic Basis of Australian Defence Policy, January 1962, in Freuling, p. 293. 223 ibid, p. 294. 224 Defence Committee’s Strategic Basis of Australian Defence Policy, October 1964, in Freuling, p. 330. 129
established. Britain had announced its withdrawal from ‘East of Suez’. Australian troops were fighting alongside US forces in South Vietnam. These developments prompted Australian defence officials to recommend that Australia develop more flexible, independent military capabilities without prejudicing Australian forces’ ability to operate with allies. Specifically, Australian forces were expected to make full use of allies’ logistic facilities and, to that end, to retain reasonable compatibility of weapons and equipment with those allies.225 While officials acknowledged the desirability of local production of equipment for Australian forces, the priority was for local capacity to maintain, repair and test equipment.226
These policy settings provided for local production of high usage items (for example, ordnance) and indigenous innovation on a niche basis (for example, sonar equipment suited to Australian warm shallow waters). Australian defence scientists had begun extending the networks they had established with their UK counterparts to include, under ANZUS auspices, contact with like-minded US scientists. In general, however, neither Australian defence policy nor Australian defence procurement gave local companies any incentive to invest in the competencies required to provide innovative solutions to Australian materiel requirements.
In July 1969 US President Nixon announced his Guam doctrine. This and South East Asia’s more favourable security outlook prompted Australian defence policymakers to move beyond the transitional emphasis on a capability for independent action and to focus on defence self-reliance. In Australian usage, this concept came to mean maintaining the military capabilities required to defend Australia and its direct interests from credible threats without relying on combat support from other countries.227 Defence self-reliance formalised the evolution that had occurred in Australia’s perceptions of its relationship with the US:
It is not our policy, nor would it be prudent, to rely upon US combat help in all circumstances. Indeed it is possible to envisage a range of situations in which the threshold of direct US combat involvement could be quite high.228
Under defence self-reliance, Australian defence planning placed more emphasis on developing and maintaining a capacity for unilateral military action in Australia’s region of primary strategic concern but within the framework of the ANZUS alliance. This shift enhanced the Treaty’s significance to Australia in terms of facilitating Australian access to advanced US technology (and intelligence). By the early 1980s, the quest for defence self- reliance had begun influencing the development of Australian defence force structure and
225 Defence Committee’s Strategic Basis of Australian Defence Policy, August 1968, in Freuling, p. 386. 226 ibid, p. 389. 227 R. Brabin-Smith, The Heartland of Australia’s Defence Policies, Working Paper No 396, Strategic and Defence Studies Centre, Canberra, 2005, pp. 2-4. 228 D. Killen, Australian Defence, Australian Government Publishing Service, Canberra, 1976, p. 10, para 8. 130
the procurement of associated defence capital equipment. Self-reliance encouraged Australian defence procurers to accord more value to indigenous ‘make’ solutions to those requirements for Australian military capability most directly relevant to the defence of Australia and its immediate interests. Self-reliance did not mean self-sufficiency in the supply and support of materiel; it meant developing the capacity to maintain, repair, modify and adapt defence equipment to the Australian environment independently of overseas sources. In order to foster such capacity, the government’s policy provided for Australian Industry Involvement (AII) in design, development and production of materiel for the Australian forces on a selective basis.229
In retrospect, the 1980s can be seen as the ‘decade of self reliance’. During this period the policy influenced such procurements as, for example, the Collins-class submarines and the JORN. The quest for self-reliant defence of Australia caused Australian defence planners to accord higher value to a capability for effective surveillance of Australia’s northern maritime approaches. JORN was the outcome of demand by Australian defence planners for an indigenous ‘make’ solution to this requirement.
During the 1980s, the government envisaged implementing AII policy through an Australian defence industrial base comprising a few large companies with the technical, financial and managerial strength to prime major defence contracts, supported by a strong layer of smaller, specialised companies able to sub-contract in a range of projects. The government envisaged both prime and sub-contractors forming strong collaborative and mutually viable relationships with overseas defence companies in order to gain access to technologies and markets.230
By the early 1990s, however, the deficiencies in the Australian defence competence bloc were becoming apparent. The symptoms included, for example, budget overruns, schedule delays and performance deficiencies in the Collins-class submarines and JORN projects. At the same time developments in the Middle East – notably the first and second Gulf Wars – led Australian defence actors to place more emphasis on expeditionary operations and less store on self-reliance. During the 1990s, these trends were reflected in the lower value accorded by defence policymakers to indigenous ‘make’ solutions to capability requirements and in their commensurately reduced appetite for the cost, schedule and technical risk inherent in indigenous military technological innovation. By the early 2000s, off-the-shelf solutions had become the default.231
229 Kim Beazley, The Defence of Australia 1987, Australian Government Publishing Service, Canberra, 1987, pp. 76-78. 230 M. McIntosh, A new environment for industry, Defence Industry and Aerospace Report, Vol 9 (11), June 1990, p. 4. 231 M. Kinnaird, L. Early and B. Schofield: Defence Procurement Review 2003, Canberra, 2003, p. 19, available at http://www.defence.gov.au/publications/dpr180903.pdf, accessed 18 March 2014. 131
Australia developed novel solutions to Australian defence capability requirements both before and after the 1980s ‘decade of self-reliance’. This thesis has focused on Australian defence innovation during the 1980s, not because those innovations were typical but because of what they reveal about the structure and operation of the Australian defence sectoral innovation system more generally.
During the Cold War, Australian defence procurement and, by extension, Australian defence innovation outcomes were influenced by Australian government procurement policies based on the principle of open and effective competition for government business. Australian government officials responsible for procurement had to satisfy themselves that the proposed procurement would achieve value for money. Australian government procurement policies hinged on the proposition that value for money was best achieved through competitive and non-discriminatory procurement processes.232 The conviction that open and effective competition among suppliers was the most effective way of obtaining, and being seen to obtain, best value for money shaped the conduct of Australian public procurement in general, and Australian defence procurement in particular.
In Australian usage open competition meant that no competent supplier was precluded from consideration for government business and that candidate suppliers were to submit tenders or proposals to procurement agencies acting at arm’s length in an accountable and transparent manner. By contrast, effective competition recognised that market circumstances might limit competition for government business and that when this occurred it might be reasonable to restrict competition for that business to selected suppliers or even to sole source the business, but that any such restriction must be made by a duly authorised government official acting in an accountable and transparent manner.
The Commonwealth Government’s commitment to procurement on a non-discriminatory basis meant that, in principle, Australian defence procurement officials were required to rank potential suppliers in terms of their commercial, legal, technical and financial abilities. They were not to discriminate against potential suppliers in terms of, for example, their size, degree of foreign affiliation or ownership, location or the origin of their goods and services.233 In practice, this principle of non-discriminatory procurement was subject to security and defence industry policy considerations. During the Cold War, however, such considerations were not allowed to diminish competition for Australian defence business:
It is this Government’s intention that, unless there are compelling reasons to the contrary, defence work will be allocated on a competitive basis using fixed price (as
232 For a current statement of this enduring principle see Department of Finance and Deregulation (Financial Management Group), Commonwealth Procurement Rules: Achieving Value for Money, Commonwealth of Australia, Canberra, 2012, p. 14, available at http://www.finance.gov.au/sites/default/files/cpr_commonwealth_procurement_rules_july_2012.pdf, accessed November 2013. 233 Department of Finance and Deregulation, p. 17. 132
opposed to cost-plus) contracts, with payments against milestones (rather than elapsed time) and with other incentives for improved performance where appropriate.234
During the Cold War, Australian governments rarely found compelling reasons to depart from the above policy. Hence this norm influenced the way Australian actors in the customer element of the defence competence bloc executed demand for solutions to Australian requirements for military capability.
7.2: Australian actors, networks and the Australian competence bloc This section focuses on the role played by Australian defence actors in performing Eliasson’s competence bloc functions as part of the wider Australian defence sectoral innovation system. In accordance with Eliasson’s model, this section describes how actors performed the informed customer function which, in the Australian defence context, is located in the public sector. This is followed by a description of how actors performed the innovator function which drives military technological innovation outcomes by addressing requirements for military capability through a combination of old and new technologies. Then follows a description of how actors performed the entrepreneur function by identifying an opportunity to meet a requirement for military capabilities through innovative technological combinations and then acting on that opportunity by marshalling the resources required to realise it. The discussion of entrepreneurs leads into a description of how actors performed the venture capitalist function by financing early start-ups and judging, ex ante, which entrepreneurs are most likely to succeed in realising an opportunity to meet a requirement for military capability. The section concludes with a description of how actors performed the industrialist function by undertaking the design, development, production, marketing and distribution required to realise the solution conceived by the entrepreneur and facilitated by the venture capitalist.
During the Cold War period, of prime interest to this thesis, the Australian defence customer was a composite located in the executive branch of the Australian system of government. The Australian defence customer comprised a Minister for Defence who was an elected Member of Parliament and who, as a member of the executive formed by the political party with a parliamentary majority, was responsible to Parliament for the Defence portfolio. In managing the Defence portfolio, that Minister was assisted by a series of non-elected experts who advised the Minister and who then implemented the Minister’s decisions. Those experts included planners who advised the Minister about requirements for new military capabilities and about the specifications of the artefacts needed to meet those requirements. The Minister was also advised by procurers, who searched for solutions to the military capability requirement, who selected solutions from the candidates identified in the search process and who procured the solution so selected. Australian defence planners and procurers were supported by enablers, who provided the specialist expertise (including,
234 Beazley, p. 82. 133
notably, scientific advice) they needed in formulating requirements for military capability, procuring solutions to those requirements, and in learning-by-using the artefacts so procured. The final component of the customer element comprised military users who were responsible to the Minister for embedding the artefacts procured to meet military capability requirements into the appropriate socio-technical regime and then learning how to use those artefacts to generate the desired military effect.
Australian innovation outcomes were influenced by, firstly, the way actors in the customer element of the competence bloc interacted with each other and by, secondly, the way the customer element interacted with other elements of the Australian defence competence bloc. The way the actors in the customer element of that competence bloc interact with each other is affected by the stringency with which the Parliament holds the Minister to account for the proper use of public funds allocated by Parliament for the conduct of defence business. The Australian Parliament is free to investigate individual items of business in as much detail as it sees fit. Such investigative scrutiny is undertaken by Parliamentary Committees who are supported by the Australian National Audit Office (ANAO). The latter is an agency of the Parliament, is independent of the executive, is legally entitled to full access to the records of executive agencies (including Defence) and is able to augment its considerable in-house expertise by engaging subject matter experts. In principle, Australian Parliamentary Committees can require Ministers to explain, on the public record, the policies and activities for which they are responsible and can require the officials advising Ministers to explain, again on the public record, the detailed decisions and choices made in executing those policies and activities. Such policies and activities include choices relating to military technological innovation.
The Commonwealth Public Service Act defines the relationship between Australian Ministers and their officials in hierarchical terms: the heads of public service agencies are responsible for managing agency business under the agency Minister; they are responsible for advising the agency Minister in matters relating to the agency and must assist the agency Minister to fulfil the agency Minister’s accountability obligations to Parliament and to provide factual information as required by Parliament.235 In implementing the Minister’s decisions, officials exercise authority delegated to them by that Minister. The exercise of such delegated authority by officials is codified in detail. Depending on the value and sensitivity of the decision involved, Australian officials responsible for formulating requirements for military capability deciding to pursue novel solutions to those requirements and the choosing between ‘make’ and ‘buy’ solutions act on the basis of a Ministerial decision or on the basis of delegated Ministerial authority.
For much of the Cold War the way the actors in the customer element of the Australian defence competence bloc interacted with each other was in continuous flux. In 1974 the
235 Public Service Act 1999, section 57. 134
Australian government accepted recommendations by Sir Arthur Tange, the then Secretary of Defence, that it synthesise into a single government department answering to the Minister for Defence on a range of defence-related activities previously dispersed among five government departments, each with their own Minister. Tange’s recommendations were driven by recognition, discussed above, that Australia needed to develop more independent military capabilities. The Tange reforms were based on the proposition that a single integrated department was the best way to enable the Minister to discharge his/her responsibilities to Parliament, firstly, for the proper management of resources appropriated by Parliament for this process and, secondly, for the exercise of command of the armed forces in accordance with Australian strategic interests.
A distinctive feature of the Tange reforms was the establishment of a formal diarchy comprising a civilian Secretary of Defence and a military Chief of the Defence Force (CDF), both answering to the elected Minister. The Secretary was responsible to that Minister for the proper management of resources while the CDF was the Minister’s chief military adviser. The development of Australian military capability, including both the formulation of military capability requirements and the allocation of the requisite resources engaged the responsibilities of both members of the diarchy. But from the inception of the Tange reforms in the early 1970s until well after the end of the Cold War in the 1990s it was the civilian organisation answering to the Secretary that dominated the preparation of strategic guidance and the adjudication of competition for funding of investment in new capabilities.
In competence bloc terms, this meant that civilian officials performed the planning function in the customer element of Australian defence competence bloc. The Australian armed forces resented this civilian pre-eminence as unwarranted intrusion by civilian policymakers into military affairs. The resulting civil-military disputation debilitated defence planning and undermined the ability of the customer element of the Australian defence competence bloc to drive Australian defence innovation outcomes from the late 1970s until nearly the end of the Cold War.236
The Tange reforms also affected the performance of the enabling function within the customer element of the defence competence bloc. During the Cold War, Australian defence actors in the customer element of the competence bloc relied almost exclusively on Australian government laboratories for local defence scientific advice. Tange consolidated these laboratories into the Defence Science and Technology Organisation (DSTO). This created a new actor within the customer element of the Australian defence competence bloc with specialist responsibilities for defence research, development, trials and evaluation. During the Cold War, DSTO undertook not only upstream scientific research but also downstream technological development. The latter activity extended to the technology
236 P. Dibb, Review of Australia’s Defence Capabilities, Australian Government Publishing Service, Canberra, 1986, p. vi. 135
demonstrator and even prototype stages. In competence bloc terms, DSTO combined both scientific enabler and innovator functions. The following paragraph discusses DSTO performance of the scientific enabler function in the Australian defence competence bloc. How DSTO performed the innovator function in the Australian defence competence bloc is discussed later in the section.
As the scientific enabler in the customer element of the Australian defence competence bloc, DSTO provided scientific advice to customer elements responsible for capability planning, procuring and using functions. DSTO complemented such domestic networks by forming bilateral and multilateral networks with comparable defence scientific organisations in other like-minded countries. Such networks comprised formally constituted forums as well as informal contacts. The forums constituted network ‘nodes’ which created both opportunities and incentives for individual scientists to make informal contact. Such informal DSTO networking was critical to JORN innovation.
Fostering bilateral networks with US agencies was a key aspect of DSTO’s enabling function. When such bilateral networks focused on a particular research activity or technological development, they were typically underpinned by government-to-government memoranda of understanding (MOU). These understandings legitimised the exchange of information, goods and services by invoking the ANZUS Treaty, then defining the purpose of the exchange and then setting out the principles (including security arrangements) to guide the activity of the officials involved. Being policy mediated rather than market based, such government-to-government arrangements established the ‘ground rules’ for informal exchange of uncodified information by scientists and other expert actors. For example, over- the-horizon radar (OTHR)-related liaison between DSTO and the US Advanced Research Projects Agency (ARPA) was formalised in an MOU that legitimised intensive exchange of tacit information by US and Australian experts.
Tange also centralised defence capital equipment procurement in a newly constituted actor, the Defence Acquisition Organisation (DAO). The intent was to secure better value for defence procurement spending by capturing DAO officials’ procurement-related learning by doing and facilitating the use of procurement to foster in local industry the capacity to supply and support materiel needed for the self-reliant defence of Australia. The DAO influenced Australian defence innovation outcomes by conducting the search for, selection of, and procurement of, solutions to military capability requirements formulated by the planners discussed earlier in this section. In conducting the search for such solutions, the DAO affected innovation outcomes by setting the parameters governing who could participate in the competition for defence business and on what terms. In selecting solutions, the DAO influenced innovation outcomes by judging the relative value for money of candidate solutions to military capability requirements, including the relative merits of ‘make’ and ‘buy’ solutions to those requirements. Finally, in procuring the solutions so
136
selected, the DAO influenced innovation outcomes by using its monopsonist status to determine and impose the terms and conditions governing the transactions involved, including, notably, which party assumed what elements of the risk associated with those transactions.
The DAO did not function autonomously. It was horizontally linked to capability planners and military users via the processes of formulating the five-year capital equipment procurement plan, of gaining ministerial approval of individual procurements and of allocating funds approved in the annual budget cycle. The DAO was vertically linked to the Minister for Defence, to whom it provided procurement advice and on whose behalf it implemented procurement decisions.
The user element of the Australian customer function comprises the Australian Navy, Army and Air Force. As users of the artefacts designed, developed and procured to meet demand for military capabilities, the three armed services are custodians of the operational knowledge accumulated through learning-by-using those artefacts. This thesis concentrates on Royal Australian Air Force (RAAF) learning-by-using radar for broad area surveillance.
Such RAAF learning-by-using had three implications for the operation of the Australian defence sectoral innovation system. Firstly, the extent to which RAAF learning-by-using influenced ongoing development of radar-related artefacts after their acceptance into military service was subject to the quality of the networks between RAAF user and artefact producer. Secondly, such learning-by-using fostered strong technological paradigms in the RAAF. For as long as those paradigms accorded with successful outcomes in the competition for military advantage, the RAAF favoured path-dependent development of the technological systems involved. Thirdly, RAAF technological paradigms were also established through learning from, and emulation of, successful use of an artefact by a peer service. In the RAAF’s case, such learning and emulation were fostered by close and long-standing cooperation with its US, UK, Canadian and New Zealand counterparts.237 Particularly relevant for the purpose of this thesis is RAAF participation in the Air Standardisation Coordinating Committee (ASCC) and its successor, the Air and Space Interoperability Council (ASIC). 238 During the Cold War, RAAF participation in the Command, Control and Intelligence, Surveillance and Reconnaissance Working Group of the ASCC and the ASIC helped foster the RAAF’s strong preference, based on United States Air Force (USAF) and United States Navy (USN) practice, for manned airborne early warning and control aircraft in meeting Australia’s demand for a broad area surveillance capability. In gaining acceptance
237 Thomas-Durrell Young, Cooperative diffusion through cultural similarity: the post-war Anglo-Saxon experience in Emily Goldman and Leslie Eliason, pp .98-99. 238 For a useful overview of these organisations see various editions of the Washington Staff Handbook and its successor, the Washington-based Military Multifora Staff Handbook, available at http://climateviewer.com/assets/img/wiretapped/2012-Multifora-Handbook.pdf. accessed 2 December 2013. 137
as a viable solution to Australia’s requirement for broad area surveillance, OTHR had to compete with this RAAF paradigm.
Australian actors performed the innovator function of the Australian military technological competence bloc by combining old and new technologies in addressing requirements for military capability. During the Cold War, this function was performed primarily by DSTO which, as discussed above, undertook the development of technology demonstrators and even prototypes. In doing so, DSTO blurred the distinction between enabler and innovator. Such role ambiguity created a dilemma for DSTO and sowed the seeds of a long-term problem for the Australian defence sectoral innovation system.
DSTO’s dilemma became apparent in its development of the OTHR technology demonstrators (Jindalee Stage A). DSTO development and production of Jindalee Stage A was critical to gaining the support of risk-averse defence planners in the contested policy environment that characterised the Australian defence organisation at the time. But the way DSTO performed this activity did little to foster an indigenous OTHR technology base outside DSTO. DSTO sharpened its dilemma in the way it developed the prototype OTHR in Jindalee Stage B. Again, successful prototyping was critical to demonstrating the potential military utility of OTHR in the Australian context. But the way DSTO managed the prototyping did little to foster OTHR-related capacity in the Australian companies who would be called on to produce the operational system.
During the Cold War, Australian defence innovation was directly affected by DSTO involvement in multilateral networks for the exchange of scientific and technical information of common interest, including, notably, The Technical Cooperation Program (TTCP). The US and the UK originally established this multifaceted forum in response to the launch of the Soviet Sputnik satellite in October 1957, after which it expanded to include Canada, Australia and New Zealand. 239 Australia joined the TTCP in 1965 and DSTO participation in the TTCP multilateral network was critical in the early development of DSTO’s competence in OTHR technology. DSTO’s deep in-house competence combined with DSTO’s overseas networks enabled DSTO to make remarkable progress at the scientific/invention end of the innovation spectrum. During the Cold War, however, DSTO’s main contact with local companies was through licensing them to market DSTO developments. Licensing DSTO inventions and developments provided little incentive for local firms to invest in the capacity to transform DSTO inventions and insights into innovative solutions to Australian requirements for military capability. This had adverse implications for the time taken to develop JORN and the cost incurred in doing so.
In the Australian defence sectoral innovation system, the quest for novel solutions to requirements for military capability encouraged the emergence of capability entrepreneurs
239 The Technical Cooperation Program, Overview, available at http://www.acq.osd.mil/ttcp/overview/ accessed 2 December 2013. 138
in the customer element of the Australian defence competence bloc and the more familiar commercial entrepreneurs in the firms populating the supplier element of that competence bloc. Within the customer element of the Australian defence competence bloc, capability entrepreneurs emerged among actors responsible for both planning and procurement functions. The shift away from forward defence to defence self-reliance gave capability entrepreneurs a strong incentive to develop the combination of codified and tacit knowledge required to judge, ex ante, what combination of technologies seemed most likely to meet a requirement for new military capability. The vague and ambivalent guidance concerning ‘make’ versus ‘buy’ solutions meant, however, that capability entrepreneurs in the planning and procurement organisations responded to such incentives in divergent ways.
Australian capability entrepreneurs were well networked with their US and UK counterparts particularly. In the Australian context, however, the benefits accruing to capability entrepreneurs through such networking were offset by the paucity of opportunities they had to learn by using that knowledge in devising novel solutions to Australian military capability requirements. In particular, opportunities for Australian capability entrepreneurs to learn from procuring a succession of novel ‘make’ solutions to capability requirements were constrained by the tendency for Australian decision-makers to choose between ‘make’ and ‘buy’ decisions on a case-by-case basis. Limited opportunities to learn meant that the business mistakes made by capability entrepreneurs in procuring ‘make’ solutions took time to rectify, in turn attenuating the schedule for devising novel solutions to military capabilities and increasing the cost of doing so.
Australian capability entrepreneurs were also required to judge which commercial suppliers offered best value for money, what procurement arrangements best aligned the interests and incentives of customer and supplier, which stakeholders in the customer element needed to be persuaded and in what terms. Actors exercising these entrepreneurial functions were mainly located in the DAO. Again, Australia’s predilection for overseas ‘buy’ solutions to its requirement for military capability limited the opportunities for Australian capability entrepreneurs to accumulate the kind of knowledge they needed to make informed judgements about the indigenous development of new technology. This caused them to make business mistakes, rectification of which took extra time and incurred extra costs.
On the supply side of the innovation system, the commercial actors (firms) involved in designing, developing and producing solutions to the demand for military capability need commercial entrepreneurs with the combination of codified and tacit knowledge and networking skill required to judge, ex ante, what combination of technologies is most likely to meet the demand for novel solutions on a profitable basis. In Australia, the ability of commercial entrepreneurs to make these judgements depended not only on their
139
commercial competence but also on the fidelity with which capability entrepreneurs communicated information about the military capability required and the incentives they created in pursuing demand for a solution to that requirement. In Australia, the paucity of innovation opportunities combined with perverse incentives resulting from misaligned institutional arrangements caused Australian commercial entrepreneurs to make business mistakes, rectification of which took extra time and incurred extra costs.
During the Cold War the Australian defence sectoral innovation system had no counterpart to Sweden’s Wallenberg as a venture capitalist. Then (as now), Australian commercial actors funded start-ups largely from retained earnings or conventional loans. However, actors in the DAO and in DSTO were able to make a modest contribution to costs incurred by commercial entrepreneurs developing defence-related technologies. In doing so, those actors performed that part of the venture capitalist function related to start-up funding. In the DAO’s case, actors responsible for fostering local industry capacity to supply and support the Australian Defence Forces (ADF) were delegated authority to make modest grants from the Defence Industry Development (DID) Fund to companies meeting strict criteria. For example, DAO officials used DID funding to assist an Australian company to develop digitally controlled valve-grinding equipment to manufacture high-precision valves for F111 aircraft hydraulic systems. In DSTO’s case, actors were delegated authority to make modest grants from the Capability and Technology Demonstrator (CTD) Fund to help local companies develop and demonstrate innovative solutions to endorsed requirements for military capability. For example, DSTO officials used CTD funding to assist an Australian company prepare DSTO-developed glide-bomb technology for production on a commercial basis.
Much development of OTHR in Australia took place in a protected niche, funded by defence capability planners through DSTO. In doing so, these actors performed that part of the venture capitalist function relating to the provision of finance for early start-ups. During the Cold War, however, Australian defence business norms kept defence capability planners and DSTO at arm’s length from commercial actors. This left them poorly placed to judge, ex ante, which entrepreneurs were most likely to succeed in realising the opportunity to develop OTHR as a solution to the requirement for comprehensive surveillance of Australia’s northern maritime approaches. Under Australian defence business arrangements, neither DAO officials nor commercial actors had significant involvement in the OTHR development niche. Their involvement began with post development procurement of JORN and required them to develop the requisite knowledge, skills and competencies (including through learning from procurement mistakes). As a result, procurement of JORN took longer and cost more than would otherwise have been the case.
A distinctive feature of the Australian defence competence bloc up to and during the Cold War was the role of the public sector in performing the industrialist function of translating
140
an idea or prototype (typically conceived by an entrepreneur) into an artefact that can be produced, marketed and distributed profitably. In Australia, the foundation of defence- oriented ‘industrialist’ competencies was established by munitions production in government factories and dockyards during the Second World War. Those factories and dockyards were a legacy of Australian governments’ historical antipathy to private production of munitions for profit. During the Second World War, Australian governments looked to government-owned, government-operated factories to maintain core munitions technical expertise and base-load manufacturing capacity, and relied on the private sector to augment this capacity as wartime exigencies required.240 A similar mind-set prevailed in the application of electrical and electronics technologies to military materiel. Hence, for example, during the Second World War radars were produced by the Post Master General’s Department and the Council for Scientific and Industrial Research. A handful of local companies manufactured electrical components like valves.
The legacy of preponderantly Australian government ownership of the means of producing defence goods and services persisted for most of the Cold War period until the government factories and dockyards were corporatised in 1988, subsequently rationalised and eventually privatised – a process which took a decade to complete, in 1999. Up to, and during, the Cold War, Australian government factories and dockyards were managed as government departments and staff operated under Australian public service terms and conditions.241 These terms and conditions were inimical to innovation generally and to electronics-based innovation in particular. The situation was exacerbated up to and during the Cold War by the lack of opportunities and incentives for private Australian companies to invest in the capacity to provide innovative solutions to Australian defence requirements. This was particularly the case in applying electronics technologies to military requirements: during the Cold War, and in the absence of local companies with the requisite electronics- related industrial competencies, procurement officials and prime contractors turned to imports, further reducing opportunities and incentives for local production. When the defence customer sought local producers as part of the quest for self-reliance, Australian companies had to establish the requisite industrialist competencies, which added time and cost to the innovation process.
The then Minister for Defence, the Honourable Kim Beazley announced the government’s decision to procure JORN in October 1986. At that time, the Australian defence industry base was dominated by government factories and dockyards engaged in legacy engineering activity in the aerospace, maritime, land vehicles and weapons and munitions sectors. In 1986, the major shipbuilding projects that were to transform the Australian defence industry (including the Collins-class submarines and the ANZAC frigates) were still
240 Mellor, p. 8. 241 See C. Coulthard-Clark, Breaking Free: Transforming Australia’s Defence Industry, Australian Scholarly Publishing, Melbourne, 1999, especially, pp. 43-61. 141
embryonic. Private companies were concentrated in the electronics sector but were relatively small and engaged primarily in sub-contract work or in production for offsets. Table 7.1 below classifies the main Australian defence industrialists operating in 1986 by defence industry sector, indicates their ownership and identifies their principal activities at that time. Table 7.1 does not include the numerous shopfronts maintained by overseas prime contractors (for example, Rockwell Collins and General Dynamics) in support of their marketing to the Australian defence customer.
Table 7.1 Australian defence industrialists 1986 (by defence industry sector)
Aerospace Maritime Vehicles & land Weapons & Electronics systems munitions
Government Aircraft Williamstown Naval Bandiana Workshops Munitions factories at Amalgamated Factory Dockyard (Commonwealth Footscray, St Marys, Wireless Australasia (Commonwealth (Commonwealth owned, rebuilding Mulwala (all (AWA); Andrews owned, F/A-18 owned, constructing armoured vehicles) Commonwealth Antennas (both assembly) naval surface owned, Australian owned, combatants) manufacturing radio ammunition & communications propellant) equipment, navigation equipment)
Hawker de Havilland Cockatoo Island Ordnance Factory Computer Sciences (overseas owned, Dockyard; Garden Bendigo; Small Arms Australia (CSA) F/A-18 components, Island Dockyard Factory Lithgow (Australian owned, aerospace offsets) (Commonwealth (Commonwealth software owned, refitting naval owned, naval guns, development) surface ships & army artillery, rifles, submarines) machine guns)
Rosebank Engineering North Queensland Faireys (overseas (Australian owned, Engineers & Agents owned, sensor computer numerically (Australian owned, systems) controlled machining building naval patrol for F111 aircraft boats)
GEC Marconi (overseas owned, Barra sonar buoys, Mulloka sonar)
Morris Productions (Australian owned, printed circuit boards for sonar buoys etc.)
142
7.3: Australian military doctrine This section analyses how Australia’s evolving strategic perceptions caused Australian defence actors to adjust military doctrines. The latter concept is defined as the combination, firstly, of the military effect planners judge necessary to achieve strategic objectives and, secondly, of how planners envisage employing military artefacts to generate that military effect. Australian innovation outcomes were influenced by how the customer element of the Australian defence competence bloc attempted to translate evolving military doctrines into requirements for military capability. Particularly relevant for present purposes are the military doctrines formulated in the 1970s-1980s, when Australian defence planning evolved from a quest for more independent military capabilities to a commitment to defence self- reliance within the framework of the ANZUS alliance. The doctrines formulated in this period included warning time and lead time, geographic priorities, the core force and defence in-depth. Each of these doctrines is discussed in the following paragraphs.
The shift from ‘forward defence’ to ‘defence self-reliance’ reflected, among other considerations, recognition by Australian defence actors that the threshold of US support could be quite high. The shift also reflected a judgement by those actors that Australia had, or could develop, the capabilities required to handle the kind of contingencies that were credible in the new, relatively benign Asian order.242 The military doctrines derived from this judgement were all based on notions of adjusting an existing force in response to future developments. The underlying premise of these notions was that the emergence of a military threat to Australia would require both military capability and political motivation. Australian defence planners reasoned that, in Australia’s circumstances, the development of capability and motivation would provide indicators that would warn Australia of the emergence of the threat.243 In the early 1970s, attempts to bridge the gap between this perception and the practical business of management of defence capability development led to the concept of balancing warning time and lead time. Managing warning time required investment in efficient intelligence, regular strategic reviews and a capacity for timely decisions to exploit warning time. Managing lead time required investment in a force-in-being of the size and with the skills and equipment that enabled it to expand in the direction and at the rate appropriate to the emergent threat.244
This reactive doctrine fostered path-dependent procurement. It provided little guidance to decision-makers choosing between ‘make’ and ‘buy’ options. Innovation was DSTO led and tended to focus on, for example, extending the service life of military aircraft and developing the ‘black box’ flight recorder. Such innovation had little impact on the time
242 Freuling, p. 45. 243 Killen, p. 10. 244 Defence Committee’s Strategic Basis of Australian Defence Policy, June 1973, in Freuling, p. 441. 143
taken to develop novel solutions to requirements, the cost incurred in doing so or the pattern of development after the innovation was introduced into service.
In an attempt to construct a more robust link between this expansion base logic and choices about what kind of military capabilities should comprise the force-in-being, Australian defence actors began in the early 1970s to accord more emphasis to structuring Australian forces with an eye to the nation’s geographic circumstances. On this basis, the self-reliant defence of Australia was judged to require, for example, capability for surveillance of coastal and off-shore resource zones, the archipelago to Australia’s north, the eastern Indian Ocean and the South West Pacific Ocean. Other capabilities included naval and air maritime defence; forces sufficient to repel or contain hostile landings on the mainland; air defence; an adequate defence infrastructure and communication network; a comprehensive intelligence organisation; and Industrial, scientific and technological support.245
This geographically oriented doctrine resonated with the emphasis on development of independent military capabilities that underpinned the Tange reforms in 1974. It fostered indigenous development of novel solutions to capability requirements on a niche basis. For example, sonar technologies suited to anti-submarine operations in Australia’s warm, shallow waters were developed by DSTO to prototype stage, and then licensed to local companies for production. However, while geographically based doctrines provided some guidance to capabilities were best suited to Australia’s circumstances, they provided limited guidance to how much Australia should spend on those capabilities and in what order. In an attempt to provide some guidance on how much to spend on what capabilities and what kind of trade-offs were appropriate, Australian defence actors introduced the notion of core force: This comprised:
a force able to undertake peacetime tasks, a force sufficiently versatile to deter or cope with a range of low-level contingencies which have sufficient credibility, and a force with relevant skills and equipment capable of timely expansion to deter or meet a developing situation. Capabilities related to the least conceivable contingency of major assault against Australia should command a low priority in the development of the force structure, provided the capability for expansion is not prejudiced.246
Attempts by Australian defence actors to apply core force principles led to a focus on the appropriate technology level of the Australian forces. For example, the guidance indicated that the level of military technology should, on one hand, be sufficient to permit peacetime tasks and responses to contingencies to be undertaken in a way which keeps down recurrent manpower and/or life-cycle costs but, on the other hand, give the defence force a capability edge in Australia’s neighbourhood. The technology level was also to be sufficiently advanced to enable Australia to develop the technical level of the force in a
245 ibid., p. 483. 246 Defence Committee’s Strategic Basis of Australian Defence Policy, October 1975 in Freuling, p. 537. 144
timely fashion, if and when more complex and developed weapons systems were called for. Finally, the technology level was to be compatible with, but not necessarily equal in technical advancement with, the relevant weapons systems of larger allies.247
Such guidance did nothing to inform choices between indigenous ‘make’ or overseas ‘buy’ procurement options. Application of core force principles to judgements about the nature and scale of local industry supply and support reaffirmed the priority for local repair, overhaul and modification of existing equipment. Capacity for local production of stores and equipment likely to be required in low-level contingencies was accorded priority while investment in local industrial capacity required for the supply of equipment relevant only to high-level contingencies was not warranted.248 Whatever the logical merit of such guidance, it was hopelessly vulnerable to the institutional problems encountered by the post-Tange defence organisation.
In the late 1970s and into the 1980s the service and civilian elements of the defence organisation, while not contesting the strategic guidance, were totally unable to agree what force structure priorities flowed from that guidance, let alone agree on the relative merits of ‘make’ versus ‘buy’ solutions.249 Such division within the customer element of the defence competence bloc did not preclude DSTO-led innovations to meet niche requirements sponsored by the services. For example, DSTO developed what became the Nulka anti-ship missile decoy system, which it later licensed to AWA Defence Industries, a local company. Similarly, DSTO developed Barra sonobuoys for anti-submarine warfare in Australian waters and a laser airborne depth-sounding system to expedite charting of Australian waters, both of which it licensed to local companies. In the Australian context, however, such licensed production did not encourage local companies to develop in-house R&D capacity. More generally, the commercial uncertainty created by intra-customer disputation over force structure priorities undermined whatever incentive local companies might have had to invest in the competencies to develop novel solutions to capability requirements.
The inability of the Australian Defence Force and the newly established Department of Defence to agree on even basic force structure concepts led the then Minister for Defence, the Hon. Kim Beazley, to commission, in February 1985, Paul Dibb to undertake a review of Australian defence capabilities. Dibb completed his review in 1986. He used Australia’s strategic geography (notably Australia’s northern maritime approaches) in gauging the upper bound of Australian military capability requirements and regional military capabilities (regardless of intentions) in gauging the lower bound of those requirements. Dibb’s review
247 Defence Committee’s Australian Strategic Analysis and Defence Policy Objectives, September 1976 in Freuling, pp. 619-620. 248 ibid, p. 621. 249 Eric Andrews, The Department of Defence (Volume V of the Australian Centenary History of Defence), Oxford University Press, 2001, pp. 245-247. 145
informed the government’s 1987 Defence White Paper which called for forces able to provide defence in depth. This required forces:
capable of tracking and targeting the adversary, mounting maritime and air operations in the sea and air gap to our north, capable of offensive strike and interdiction missions, having a comprehensive range of defensive capabilities – including air defence, mine countermeasures, and protection of coastal trade – and embodying mobile land forces able to defeat hostile incursions at remote locations250
Recognition of the priority for strategic control of the sea and air gap to Australia’s north was a necessary condition for investment in OTHR but not a sufficient condition. As discussed in Chapter 8, actors in the customer element of the Australian defence competence bloc held varying views as to the most appropriate way to provide the surveillance required for such control. Reconciling these divergent views not only took time but was a precondition for investment by Australian companies in competencies required to design and develop a solution to requirement for effective surveillance of the sea and air gap based on the OTHR technology developed by DSTO.
7.4: Australian Technology In searching for a solution to Australia’s requirement for a broad area surveillance capability during the Cold War, Australian defence actors concentrated on radar technology and did not attempt to emulate, for example, the US exploitation of satellite-based surveillance technology, then focused on visible light imagery and thermal infra-red imagery.251 This section analyses the Australian radar technology base in terms of, firstly, the prevailing radar technological paradigm. The latter is defined as a ‘pattern’ of solutions to selected techno-economic problems based on selected principles derived from the natural sciences, jointly with specific rules aimed at acquiring new knowledge – and safeguarding it, whenever possible, against rapid diffusion to the competitors.252 The section then analyses the radar technological systems comprising the Australian radar technology base. In this thesis a ‘technological system’ comprises a network of agents interacting in a specific economic/industrial area under a particular institutional infrastructure and involved in the generation, diffusion and utilisation of a specific technology or related group of technologies.253 The section then builds on the discussion of technological systems by analysing the trajectory of radar technological development in the Australian technology base. The ‘trajectory of technological development’ is defined as the path of improvement
250 ibid, p. 31. 251 For a full discussion of US Cold War satellite surveillance see J. Richelson, America’s Secret Eyes in Space: The US Keyhole Spy Satellite Program, Harper Collins, 1990. 252 Dosi, Sources, procedures and microeconomic effects of innovation, p. 1127. 253 B. Carlsson and R. Stankiewicz, On the nature, function and composition of technological systems, Journal of Evolutionary Economics, Vol 1, 1991, p. 93. 146
followed by a technology in response to technologists’ perceptions of opportunities and having regard to how the market and other evaluation mechanisms determine what kinds of improvement would be profitable.254
During the Cold War, the RAAF was primarily responsible for broad area surveillance of the Australian littoral. The RAAF therefore established the predominant broad area surveillance technology paradigm. For virtually the entire Cold War period, the RAAF used microwave search radars embarked on successive generations of long-range maritime patrol (LRMP) aircraft to conduct surveillance of the Australian littoral, including of the continent’s northern maritime approaches. RAAF airborne surveillance was augmented by successive generations of Royal Australian Navy (RAN) patrol boats which operated shorter range microwave surveillance radars. For the RAAF, long-range maritime patrol was an adjunct to its primary task of anti-submarine warfare, for which the RAAF LRMP aircraft were optimised. RAAF preoccupation with anti-submarine warfare and the inherent limitations of the RAN patrol boats’ capacity to provide broad area surveillance resulted in relatively perfunctory surveillance. This became progressively less acceptable as Australia began to enforce its maritime resources claims in the 1960s and as effective control of Australia’s northern maritime approaches was accorded higher strategic priority in the 1970s.
In adjusting to the demand for more effective maritime surveillance, the RAAF retained its technological paradigm defined by microwave surveillance radars and embarked on manned aircraft. However, the RAAF extended its initial focus on a capability for anti-submarine warfare operations to include a capability for airborne early warning and control (AEW&C) as part of its planning for air superiority operations in defence of northern Australia and the associated northern maritime approaches. In pursuing an AEW&C capability, the RAAF was influenced by that element of USN and USAF broad area surveillance activity based on airborne microwave radar technology. In the USN case, the main exemplar of this technology was the Grumman E-2C Hawkeye AEW&C aircraft which entered USN service in 1964 to meet the USN demand for an all-weather, carrier-based tactical airborne early warning and command and control capability for USN Carrier Strike Groups. The E2-C is a twin-engine high-wing turboprop aircraft with a distinctive 8 m diameter radar rotor dome mounted dorsally, above the wing. The E-2C’s basic operational concept is to gather surveillance data using its mechanically scanned radar and downloading that data to the Carrier Group’s command centre. As such, the E-2C Hawkeye was not well suited to the widely dispersed, low-intensity surveillance operations required in Australia’s circumstances.
In the USAF case, the main exemplar for the RAAF was the Boeing E-3 Sentry airborne warning and control system (AWACS), which entered USAF service in 1977. Like the E-2C
254 G. Dosi and R. Nelson, An introduction to evolutionary theories in economics, Journal of Evolutionary Economics, Vol 4, 1994, p. 161. 147
Hawkeye, the E-3 Sentry also carries a distinctive dorsally mounted rotating radar dome (over 9 m in diameter). Unlike the E-2C, however, the E-3 Sentry constituted a complete command post, embarking numerous operators performing surveillance, identification, weapons control, battle management and communication functions for numerous targets. The E-3 Sentry is, arguably, the world’s most capable AEW&C system but it is complex and expensive both to acquire and to operate.
During the Cold War, the element of the Australian radar technological system relevant to broad area surveillance was shaped by actors in the customer element of the Australian defence competence bloc. Within that element, the two most significant actors were the defence scientists (enablers) and the RAAF (users). During the Anglo-Australian Joint Project (1946-80) at Woomera, Australian defence scientists in what was to become DSTO used specialised radars imported from Britain and later the US to search for and track British rockets. 255 While such radar-related research activity did much to establish radar user competencies, it was largely confined to the customer and inventor elements of the Australian radar competence bloc. The RAAF’s requirement for microwave search radars was met by leveraging the ANZUS relationship and importing relatively advanced radars from the US. The imported radars were either repaired, maintained and upgraded in-house by RAAF technical personnel or returned to the US. Local innovations that did occur were confined to embedding ‘buy’ solutions into the RAAF or RAN socio-technical regimes. This pattern of procurement and use created a cycle of causation. The customer element of Australia’s radar competence bloc could meet its needs by accessing overseas radar technology. Australian firms like AWA and Fairey, who might have fostered commercial actors able and willing to perform the innovator, entrepreneur and industrialist functions of a radar competence bloc, had neither opportunity nor incentive to establish the requisite competencies. Lack of relevant local innovator, entrepreneur and industrialist competencies in radar technological systems reinforced the customer’s predilection for overseas solutions.
Microwave radar signals are not refracted by the ionosphere and, hence, are not suitable for over the horizon applications. By contrast, the ionosphere will refract high frequency (HF) radar signals (subject to the angle of incidence of the signal to the ionosphere). This phenomenon can be used for detecting targets over the horizon, beyond line of sight. The British Chain Home radar that diffused to Australia in 1939 operated in the HF band. During the Second World War, however, military demand for greater precision and sensitivity and Britain’s invention of the magnetron in 1940 meant that HF radar was rapidly superseded by progressively more capable microwave radars, which also diffused rapidly to Australia.256
By the end of World War Two microwave radar technology had largely eclipsed HF radar technology in both civil and military applications in Australia and elsewhere. Specifically, by
255 P. Morton, Fire Across the Desert – Woomera and the Anglo-Australian Joint Project 1946-1980, Australian Government Publishing Service, Canberra, pp. 294-297. 256 Zimmerman, pp. 107-129. 148
the end of World War Two, microwave radar had become the foundation of the RAAF radar technological paradigm. Hence the HF radar that was to be the vehicle for OTHR development leading to JORN was virtually eliminated from the Australian radar technology base and had to be re-established. It was by pure happenstance that Strath was using an obsolescent British HF radar to track British rockets as part of the Joint Project when he observed the ionospheric refraction of that radar frequency. It was also serendipitous that W.S. Butement (the Chief Scientist in the Department of Supply with overall responsibility for the provision of scientific support for the Joint Project) had participated in Britain’s Chain Home radar development and in other World War Two radar programs and, as a result, recognised the potential of Strath’s OTHR work and supported further experiments.257
The US Naval Research Laboratory had begun experimenting with HF-based OTHR in the 1950s and, by 1956, had succeeded in using ionospherically refracted HF signals to track aircraft flying across the Atlantic. The USN and USAF subsequently invested in OTHR systems, primarily for defence of the continental US. The US Defence Department’s Advanced Research Projects Agency granted Strath and his DSTO colleagues access to the results of US experiments in OTHR and provided considerable practical support, thereby expediting the Australian work.
But an OTHR-based solution to Australia’s broad area surveillance requirement was completely at odds with the prevailing Australian microwave radar technology paradigm. Australian OTHR advocates had to displace this predominant technological paradigm in order to establish an OTHR-based technological system. DSTO access to US research into OTHR enabled DSTO to demonstrate OTHR’s military utility relatively cheaply and quickly. However, much more time and money was required to displace the established Australian microwave radar paradigm sufficiently to convince the customer element of the Australian defence competence block to invest in an OTHR-based broad area surveillance system. At least initially, the RAAF viewed proposals for the use of ground-based OTHR technology for broad area surveillance as not only inconsistent with its airborne microwave radar paradigm (and therefore to be contested), but also as inimical to timely acquisition of airborne early warning aircraft (and therefore to be resisted).258 In these circumstances, funding for Australian OTHR development depended on support by Australian defence planners. By the 1980s the latter acknowledged the fundamental importance of effective surveillance of Australia’s northern maritime approaches to Australian defence self-reliance. But civilian planners in particular regarded the highly capable but expensive AEW&C aircraft advocated by the RAAF as both unaffordable and inconsistent with Australia’s relatively benign strategic outlook. This position was supported by the Australian Navy who was concerned
257 D.H. Sinnott, The Development of Over-the-Horizon Radar in Australia, DSTO Bicentennial History Series, Canberra, 1988, p. 8, available at http://www.dsto.defence.gov.au/attachments/The_development_of_over-the- horizon_radar.pdf, accessed 20 April 2012. 258 See, for example, Sinnott, The Development of Over-the-Horizon Radar in Australia, Chapter 9, p, 3. 149
that funding AEW&C aircraft would prejudice funding for frigates. The resulting disputation within the customer element of the Australian radar competence bloc directly affected the time taken to develop JORN and the cost of doing so. Nevertheless, DSTO did succeed in building an OTHR technology demonstrator by 1979 and, subsequently, an OTHR prototype that was operational by the early 1980s. Knowledge of these developments, however, was confined to Australian defence scientists working at the invention end of the spectrum of innovation activity. The insights were not shared with any Australian companies who, consequently, had neither incentive nor opportunity to develop the competencies required to exploit the insights in responding to defence or other needs.
By the early 1980s, then, the Australian radar technology base comprised the dominant microwave radar technological system and a nascent HF radar technological system for OTHR purposes. A brief description of the different development trajectories followed by these two radar technological systems helps explain JORN innovation outcomes. Australia was a fast follower of military microwave radar innovation, particularly that undertaken by the US. In the case of radar, Australia typically acquired such innovations embedded in platforms procured from the US. For example, in 1981 Australia procured 75 McDonnell Douglas F/A-18 aircraft, fitted with (then) state-of-the-art AN-APG 65 radars (manufactured by Hughes Corporation in the US) and associated electronic countermeasures (ECM) systems. While the US Government was prepared to release the physical radar and ECM systems to Australia, it was not prepared to release the software algorithms that Australia required to program the radar and ECM systems for fully sovereign operation of the aircraft. These systems were returned to the US for repair, maintenance and adaptation.259 As the then Minister for Defence, the Hon Kim Beazley subsequently explained, Australia’s inability to reprogram the AN-APG 65 radar and associated ECM systems (which were configured for USN operations against the Warsaw Pact forces) meant that Australia’s F/A-18 could not identify the US and other Western combat aircraft being procured by Australia’s neighbours.260
This lacunae in the Australian microwave radar technology system had serious implications for the credibility of Australia’s efforts to provide for the self-reliant defence of Australia. As part of a concerted effort to remedy the situation, Australia eventually obtained a substitute radar and attempted (unsuccessfully) to develop an indigenous substitute for the ECM systems.261 The difficulties Australia had with the F/A-18 fighter’s operationally critical radar and ECM systems influenced development of the nascent HF-based OTHR technology system. When a decision was made to procure JORN in 1986, actors in the customer element of the Australian defence competence bloc were determined to avoid a repetition
259 Industry Involvement and Contracting Division, Department of Defence, Review of the F/A-18 Industry Program, March 1994, Annex B, Table 3 F/A-18 Repair Venues By Major System. 260 K. Beazley, (Member for Brand), Hansard. House of Representatives, Canberra, Thursday 20 September 2007, pp. 37-44, available at http://www.aph.gov.au/Hansard/reps/daily/dr200907.pdf accessed 19 May 2009. 261 P. Hall and R. Wylie, Arms export controls and the proliferation of military technology, pp. 62-66. 150
of the above F/A-18 situation. Despite generous US support for DSTO’s work on OTHR, these actors were concerned to establish and maintain full control of Australian-developed OTHR technology. The practical consequence of this was a concerted effort by those actors to develop an indigenous HF-based OTHR technology system.
7.5 Australian demand During the Cold War, demand for solutions to Australian capability requirements was executed by the customer element of the Australian defence competence bloc within the framework of the norm of open and effective competition. This section analyses how Australian defence innovation outcomes were influenced by the way the Australian defence customer searched for a solution to a requirements for Australian military capability, selected a solution to that requirement and procured the solution so selected.
The tendering process provided the institutional connection between, on one hand, the norm of open and effective competition and, on the other hand, the search by the Australian defence customer for novel solutions to military capability requirements. Within the customer element of the Australian defence competence bloc, the DAO could search for both solutions and suppliers of those solutions via open tender processes, prequalified tender processes, limited tender processes and, by extension, the single supplier limited tender process.262
Under the open tender process, officials published an open approach to the market inviting submissions from all interested potential suppliers. During the Cold War, stakeholders in Australian defence procurement generally recognised that an open tender process initiated via a Request for Tender (RFT) provided the most competitive and non-discriminatory search process. DAO officials used the open tender process when they sought to maximise competition for the business concerned, or when they did not know the market. During the Cold War, however, the Australian defence customer generally recognised that open tendering was unlikely to yield good value for money where the defence customer accorded high value to the effectiveness with which the artefact procured performed a military function but was unsure as to which technology was best suited to produce such an artefact or what was an appropriate price to pay for that artefact. Because this was the case with JORN, the Australian defence customer eschewed an open tender process.
During the Cold War, the Australian defence customer attempted to reduce the transaction costs incurred in tendering for innovative solutions to military requirements by using the prequalified tender process. The most basic variant of the prequalified tender process entailed selecting a supplier from a multi-use list (MUL). The latter is a list of prequalified suppliers who have responded to an open approach to the market seeking to identify
262 Defence Materiel Organisation, Defence Procurement Policy Manual: Mandatory Procurement Guidance for Defence and DMO Staff, Department of Defence, Canberra, 2013, pp. 3.1-4 to 3.1-11 available at http://www.defence.gov.au/dmo/gc/dppm.cfm, accessed November 2013. 151
suppliers eligible and interested in supplying specified goods and services under specified terms and conditions and who have satisfied eligibility criteria for inclusion on the MUL. Another variant of the prequalified tender process used frequently during the Cold War entailed conducting an open approach to the market and seeking an expression of interest from potential suppliers or inviting suppliers to register interest or requesting a proposal from suppliers. Such approaches to the market included relevant requirements and evaluation criteria. Australian defence procurers used this process to establish a shortlist of potential JORN suppliers by using the responses they received to assess, ex ante, the extent to which a respondent was likely to meets the JORN performance specifications.
During the Cold War, Australian defence officials were able to dispense with the usual open approach to the market and invite either a single potential supplier or a number of potential suppliers to submit a response. Such limited tendering was only permitted in specifically defined circumstances and then only after the officials had justified doing so in writing. Circumstances warranting a limited tendering approach included the existence of only a few known suppliers able to supply the required goods and services. A limited tendering approach was also acceptable when technical considerations precluded competition, when the requirement was genuinely urgent, where only a few known potential suppliers were able to meet Defence security requirements or where the cost of an open tender process would be so expensive as to preclude value for money.
When Australian Defence officials undertook limited tendering ‘due to an absence of competition for technical reasons’, they were expected to demonstrate that they developed the requirements and specifications for the procurement based on a sound and unbiased understanding of market capabilities and commercial practices. They were also expected to undertake such market research necessary to develop a comprehensive knowledge of the specific market. After Australian procurement officials had used prequalified tendering processes to shortlist candidate suppliers, they then used limited tendering to select the initial JORN prime contractor. When this tenderer encountered difficulties, Australian procurement actors undertook another round of limited tendering due to an absence of competition for technical reasons. In doing so, they were required to prepare a written report (which was available to the ANAO) describing the value and type of goods or services so procured, explaining the circumstances and conditions that justified the limited tender process and demonstrating how the procurement represented value for money in the circumstances.
In gauging which solution to a requirement for military capability proffered best value for money, actors performing the Australian defence customer function were required to assess, ex ante, the relative economy with which candidate suppliers proposed to provide artefacts meeting the requirement, taking into account whole-of-life costs. The relative efficiency with which candidate suppliers could produce artefacts meeting the requirement,
152
having regard to those suppliers’ experience and record of performance, was of critical importance. Other critical factors included the relative efficiency with which candidate artefacts would meet the requirement, having regard to the cost, schedule and technical risks involved and the relative effectiveness with which candidate artefacts would meet the requirement, including not only the artefacts’ fitness for purpose but also their flexibility and their ability to adapt to evolving requirements over time.263 In making the above judgements, the decisions made by the Australian defence customer were subject to detailed scrutiny. In the JORN case, the DAO selection decisions were the subject of detailed scrutiny by the ANAO and, subsequently, by the Joint Parliamentary Committee of Public Accounts. Procurement officials expected such scrutiny and conducted themselves accordingly.
A distinctive feature of Australian defence procurement processes was the exercise of delegated ministerial authority by actors in the customer element of the defence competence bloc. In Australian governance, ministerial authority is delegated to a specific position in the department or to a class of positions in the department. The authority so delegated is limited to specific functions and to defined financial amounts and must be exercised with due diligence. Finally:
A delegate’s judgement may be open to internal or external scrutiny. Therefore a delegate must always document their decision and the basis upon which it was made, and be prepared to justify it. The written record of the delegate’s submission must be sufficiently detailed to withstand review. The level of detail will vary according to the complexity, value and overall risks associated with the relevant procurement.264
In general, the prospect of scrutiny caused Australian defence procurement officials to prefer fixed-price contracts, which required the contractor to assume the burden of risk. Towards the end of the Cold War, however, Australian procurers began using turnkey contracts. Under these arrangements, Defence specified the performance it required to meet a military capability requirement and looked to the contractor to assume full responsibility for designing, developing and building a solution that delivered the requisite performance. This method of procurement entailed greater knowledge of company capabilities and more detailed understanding of the incentive structures than previous arrangements. JORN was an early turnkey contract. The quest for self-reliance also prompted demand for more novel ‘make’ solutions that entailed a degree of cost uncertainty that precluded use of firm or variable price contracts but that did not justify resort to a cost-reimbursement contract. This prompted the DAO to experiment with target cost-incentive contracts. These constitute a variant of the traditional cost-reimbursement contract but with explicit cost controls and incentives designed to give better outcomes for
263 DoFA, Commonwealth Procurement Rules, p. 15. 264 DAO, Defence Procurement Policy Manual, p. 1.4-9. 153
Defence. The incentive component of the contract is typically based on a pain share/gain share formulae that operates to adjust the contractor’s fee according to its performance against an agreed target cost.265 JORN was an early target cost-incentive contract.
7.6 Conclusion This chapter has described the structure of the Australian defence sectoral innovation system in terms of the five ‘building blocks’ – institutions, defence competence bloc, military doctrine, technology base and execution of demand. In doing so the chapter drew attention to the distinctive features of those ‘building blocks’ that, prima facie, were likely to affect the performance of the system in terms of the time taken to generate novel solutions to Australian capability requirements, the cost incurred in doing so and the pattern of development/diffusion of those solutions. The next step is to analyse how those features actually affected the performance of the system. To this end, the next chapter analyses how the distinctive features of the Australian defence sectoral innovation system affected the time taken to develop JORN, the cost incurred in doing so and the pattern of JORN development after its acceptance into Australian service.
265 ibid, p. 2.7-5. 154
Chapter 8: The Jindalee Operational Radar Network As the next step in analysing the performance of the Australian defence sectoral innovation system, this chapter describes how it works. In order to identify the causal linkages involved in the working of the system, this chapter describes the systemic response to the requirement for effective but affordable surveillance of Australia’s northern maritime approaches in terms of the development and procurement the JORN and then its eventual integration into a portfolio of broad area surveillance assets. Accordingly, Chapter 8 begins with an overview of the development of the OTHR in Australia and the US and a brief discussion of sources of information about these developments and of JORN. This is followed by a description of the initial investigation in Australia and the US of the feasibility of using ionospheric refraction of HF radar signals for surveillance purposes. The chapter then describes the Australian development of an OTHR technology demonstrator (Project Jindalee). This description is followed by a discussion of the process by which Australian defence actors formulated the demand for an OTHR-based solution to Australia’s requirement for a broad area surveillance capability (JORN). This discussion then leads to a discussion of how Australian defence actors searched for a JORN supplier, selected that supplier and attempted to procure JORN from the selected supplier, only to replace the initial supplier with another. The chapter concludes with a discussion JORN’s development after it was accepted into service with the RAAF.
8.1 Developing JORN: Overview Australia was one of several countries that, early in the Cold War, investigated the feasibility of exploiting ionospheric refraction of HF radar signals for broad area surveillance at long range. The US also investigated the phenomenon. As Figure 8.1 indicates, while early Australian investigations faltered, concurrent US experiments succeeded in demonstrating the feasibility of using ionospherically refracted HF radar signals to track aircraft at long range in the late 1950s. As Figure 8.1 also shows, Australian exploitation of the phenomenon in developing and deploying a network of ground-based OTHRs to Australia’s requirement for a broad area surveillance capability proceeded in four stages. In the first stage Australian defence scientists used the known scientific phenomenon of ionospheric refraction of HF radar signals to develop, with US assistance, a workable OTHR system called Jindalee. The second stage entailed transferring the OTHR knowledge so gained to commercial companies contracted by the Defence Department to develop and build a series of interlinked and overlapping OTHRs called the JORN. The third stage entailed embedding the JORN system into the portfolio of air defence assets maintained by the RAAF to provide effective surveillance of Australia’s northern maritime approaches. The fourth stage entailed upgrading the JORN system by taking advantage of both knowledge gained through use of the system and ongoing technological development to improve its efficiency and effectiveness.
155
Figure 8.1 Overview of JORN development, procurement and upgrade
In order to gauge the performance of the Australian defence sectoral innovation system, this chapter provides data about the time taken to develop JORN, the cost incurred in doing so and the pattern of JORN development after the system was accepted in RAAF service. Unless otherwise indicated, the data for JORN schedule and cost is drawn from a combination of Professor Don Sinnott’s monograph on the development of OTHR in Australia,266 reports by the Australian National Audit Office267 and reports by Parliamentary Committees.268 Professor Sinnott’s monograph is the best publicly available source of early OTHR development in Australia, being based on his personal involvement in the JORN project and on his access to DSTO documentation. The ANAO reports are based on full access to Defence documentation (including correspondence with members of the JORN Project Office and other Defence officials) and are the best publicly available source of information about defence involvement in the initial JORN procurement. The Parliamentary reports are based on the Hansard records of the hearings conducted by Parliamentary Committees and on submissions to the Committees by both Defence and commercial witnesses. The Parliamentary reports complement the ANAO material and provide the best
266 D. Sinnott, The Development of Over-the-Horizon Radar In Australia, p. 6. available at http://www.dsto.defence.gov.au/attachments/The_development_of_over-the-horizon_radar.pdf. accessed 1 December 2013 267 Notably the report by R. McNally, Jindalee Operational Radar Network (Australian National Audit Office Performance Audit No 28 1995-96), Australian Government Publishing Service, Canberra, 1996. 268 Especially the report by the Joint Committee of Public Accounts and Audit, Report No 357 – The Jindalee Operational Radar Network, Commonwealth of Australia, Canberra, 1998. 156
publicly available source of information about both commercial actors involved and their interaction with Defence officials.
8.2 Initial investigation of OTHR technology This section covers the genesis of OTHR research in Australia, the early opportunistic investigation of OTHR as part of the Anglo-Australian Joint Project and the more purposeful investigation of the properties of the Australian ionosphere (Project GEEBUNG). Ionospheric refraction of radio waves enabled Marconi to transmit and receive a radio signal across the Atlantic in 1901. The international scientific community had established a reasonable understanding of the radio-physics of the phenomenon by the 1930s. At this time, Australian research was focused on the phenomenon’s implications for communications and was concentrated in Sydney University and the Radio Research Board of the Commonwealth Government’s Council for Science and Industrial Research (CSIR).
In 1949 the CSIR and its successor, the Commonwealth Scientific and Industrial Research Organisation (CSIRO), stopped military research and development for a combination of security reasons269 and philosophic objections.270 In 1950 the Commonwealth Government established the Australian Defence Scientific Service within the Research and Development Division of the Department of Supply and Development.271 After 1955, work related to military radar was undertaken by the Electronic Research Division in the Long Range Weapons Establishment of the Australian Defence Scientific Service. At about this time, scientists from the Electronic Research Division investigated the feasibility of using ionospheric refraction of HF radar signals to detect aircraft at long range. Inadequacies of the technology then available precluded worthwhile results and the work was abandoned in 1955.272
The above Australian work proceeded in parallel with, and apparently in ignorance of, highly classified work on OTHR by the US Naval Research Laboratory (NRL). NRL’s success in 1956 in tracking aircraft using ionospherically refracted HF signals was due to its use of innovative signal processing technology (called magnetic drum radar equipment – MADRE). This technology, developed by NRL in-house, enabled NRL operators to extract the weak signal reflected by the aircraft and refracted back through the ionosphere from the mass of electrical noise and background clutter. NRL operators could detect aircraft by identifying the Doppler shift caused by the moving aircraft in the frequency of the signal it reflected back to the MADRE receiver.273
269 Morton, pp. 104-108. 270 B. Collis, Fields of Discovery: Australia’s CSIRO, Allen & Unwin, Crows Nest, 2002, pp. xiii-xix. 271 P. Donovan, Anticipating Tomorrow’s Defence Needs: A Century of Australian Defence Science, Commonwealth of Australia, Canberra, 2007, pp. 37-40. 272 Sinnott, p. 6. 273 J. Headrick and M. Skolnik, Over-the-horizon radar in the HF band, Proceedings of the IEEE, Vol 62 (6), June 1974, p. 665. 157
Participation in the Anglo-Australian Joint Project (1946-80) exposed Australian defence scientists to advanced radar technology. In 1958, John Strath, an Australian defence scientist, used British-designed and -manufactured HF radar to track a British Black Knight rocket launched in a near-vertical trajectory from Woomera. As the rocket left the earth’s atmosphere its radar echo was greatly enhanced by reflection of the HF radar signal by the ionosphere. Senior Australian defence scientists commissioned Strath to further investigate the phenomenon as a possible means of detecting the launch of inter-continental ballistic missiles.274 Strath, a physicist by training, had extensive experience in the development and deployment of radar in Britain during World War Two. He had emigrated to Australia to join the fledgling Defence Scientific Service in 1952, had been exposed to the abortive work on use of ionospherically refracted HF radar signals to track aircraft, and had subsequently become involved in a series of radar experiments using rockets launched from Woomera as part of the Anglo-Australian Joint Project.275
Strath’s investigation continued during the early 1960s, using improved instrumentation to study ionospheric and Doppler effects with increasing precision. By December 1965 both Strath and his research colleague were convinced that the Australian ionosphere was sufficiently stable to enable aircraft detection at long range by Doppler processing of OTHR signals. Strath and his colleague took some seven years – a total of 21 scientist years – to reach this point through opportunistic research. Strath worked in isolation from comparable work overseas. Such opportunistic research enabled them to accumulate valuable data about the Australian ionosphere and to develop expertise in conducting the ionospheric radar measurements required to progress the idea of OTHR-based detection of aircraft. By the end of 1965 Strath had recognised that a more structured program of research entailing the sustained application of larger resources was required to pursue this idea.276
In the mid-1960s Australian defence planning was still dominated by the logic of forward defence which, as indicated in Chapter 7, was inimical to indigenous innovation. In order to engender more support for the indigenous research program he envisaged, Strath took advantage of Australia’s recent accession to the TTCP to engage the US. As a key node in the network of Anglophone defence scientists, the TTCP enabled Strath to test his ideas and to identify like-minded interlocutors with complementary competencies who could help him develop those ideas. In September 1968 (some 10 years after he began his ionospheric work), Strath presented the results of his research to a TTCP subgroup. Strath’s presentation gained him an invitation by the ARPA to visit the US to discuss then highly classified US OTHR work. ARPA supported Strath’s idea of OTHR-related collaboration between US and Australian defence scientists and, to that end, agreed to sponsor the requisite government- to-government agreement. Strath led another larger specialist mission to the US in October
274 Sinnott, The Development of OTHR in Australia, p. 8. 275 D. Sinnott, John Strath, unpublished monograph. 276 ibid, pp. 14-15. 158
1969 in order to inform his proposals for a series of collaborative studies aimed at defining Australian operational requirements and establishing the feasibility of an Australian OTHR.
The idea of using OTHR for the surveillance of Australia’s northern maritime approaches resonated with emergent thinking among defence planners about the need for more independent military capability geared to the defence of Australia. US support for such studies reassured risk-averse Australian defence policymakers that the idea of using OTHR to detect aircraft at long range had sufficient potential to warrant investigation. The NRL’s work showed the way ahead. The cost incurred and time taken to complete this opportunistic phase of OTHR development in Australia are summarised in Table 8.1
Table 8.1: Opportunistic research
Dates Activity Time Cost incurred
Elapsed (years) ($Am)
1958-Sept 1968 Opportunistic research 10 30 scientist years @
(Joint Project) $26,300/year = 0.8277
Strath finalised his proposed program of studies in July 1970, some four years before the Tange reorganisation of the Defence group of departments. Gaining interdepartmental support for the proposed program took Strath some four months and it was not until November 1970 that the then Minister for Defence approved the first phase of the studies, designated Project Geebung. This entailed the commitment of some 19 scientist years of effort over a 16-month period at an estimated cost of $0.5 million. Project Geebung involved both ionospheric measurements and operations research. The measurements were intended to establish how well or otherwise the ionosphere would support OTHR operations in the Australian environment. The operations research was intended to establish the role of, and requirement for, an OTHR in the self-reliant defence of Australia.
The measurements confirmed that the Australian ionosphere was entirely compatible with OTHR operations, although more work was required to better understand its behavioural parameters. The studies also highlighted the need to establish a pilot program – what Schott and Geels would call a technological niche278 – to enable Australian scientists to refine the technology and develop the requisite competencies. Strath also considered ongoing US- Australian cooperation vital, in order to compensate for known deficiencies in the Australian scientific knowledge base and to develop OTHR at a cost and in a timeframe acceptable to
277Estimate of $26,300 for the cost of one scientist year derived from the Project Geebung budget of $0.5 million for 19 scientist years. 278 Schot and Geels: Niches in evolutionary theories of change, p. 615. 159
Australian decision-makers. Project Geebung entailed intense networking between the US and Australian OTHR actors: a US team visited Australia in February 1972 to firm up a US- Australian program of cooperative OTHR research and Strath participated in a reciprocal visit to the US, also in 1972, to refine Australian plans for post-Geebung development. The upshot was plans for a phased program of post-Geebung development of OTHR, involving progressive escalation in complexity, capability and cost.279 In accordance with the schedule approved by the Minister, Project Geebung lasted two years, being wound up at the end of 1972.
The cost incurred and time taken to complete the Project Geebung phase of OTHR development in Australia are summarised in Table 8.2.
Table 8.2: Project Geebung
Dates Activity Time Cost incurred ($Am) Elapsed (years)
Nov 1970 – end Project Geebung 2 0.5 1970 (ionospheric measurement)
8.3 Project Jindalee – Developing OTHR for broad area surveillance Project Jindalee was the name given to the staged activities through which Australian defence actors ascertained the utility of OTHR as a solution to Australia’s requirement for a broad area surveillance capability. Project Jindalee comprised three stages. The first, Jindalee Stage A, involved the establishment of the OTHR test bed near Alice Springs and gathering data for use in developing the signal processing software for Stage B. The second stage, Jindalee Stage B, involved the modification of the fixed, staring OTHR beam used in Stage A to a scanning beam. The third stage involved the establishment of the Jindalee Facility Alice Springs (JFAS) as a test bed for ongoing development of OTHR technology (while JORN was built) and as a resource for training Service personnel to operate JORN when it came on line.
DSTO spent most of 1973 planning Jindalee Stage A in detailed consultation with ARPA. Like the NRL’s MADRE program, Jindalee Stage A was intended to demonstrate the feasibility of using HF radar to track aircraft under varying ionospheric conditions. To this end Strath
279 Sinnott, The Development of OTHR in Australia, pp. 16-20. 160
chose a site near Alice Springs, in central Australia. The site lay under an international air corridor, was electrically quiet and logistically supportable.
Jindalee Stage A was also intended to foster an indigenous OTHR technology base while taking advantage of US support to reduce the cost incurred and time taken to do so. The technology base so established, however, was confined to DSTO: for example, DSTO scientists designed and developed both the wave-form generator (required to generate the extremely pure signals needed for managing ionospheric refraction) and the signal processor (required to extract the weak return signal from the background clutter).280 DSTO’s in-house competencies obviated the need to engage local companies in this work. In addition, the secrecy of the work led Strath to insist on a strict need-to-know regime which reinforced DSTO’s tendency to exclude local companies from substantive engagement in the research.
The then Minister for Defence approved Jindalee Stage A in April 1974 at a cost of $3.4 million over 5 years.281 Jindalee Stage A was a technology demonstrator. It used non- scanning HF radar beams refracted through the ionosphere to detect commercial aircraft flying routine schedules over north-western Australia. Scientists detected targets manually, from a display of range Doppler data. Despite its experimental nature, the program began to encounter opposition. While Jindalee’s potential was readily recognised by central defence planners, both military and civilian, OTHR had no constituency in Navy or Air Force. In an indication of the immaturity of the Tange reforms (initiated in 1974), the single Service chiefs saw OTHR development as inimical to established concepts of operation. The then Chief of Air Force in particular resented it as, in his view, a wholly unwarranted intrusion into an operational matter – broad area surveillance – for which he and RAAF air crew were properly responsible.282
In these circumstances, maintaining support for the program required the DSTO scientists to demonstrate continuously the system’s operational potential to decision-makers and stakeholders. Despite the limitations of the experimental Stage A equipment, the scientists demonstrated the system’s ability to detect an aircraft manoeuvring off Australia’s northwest coast in April 1977 and to detect ships in northwest Australian waters in December 1977. These sorts of demonstration took extra time and cost an additional $2.6 million. They did, however, enable OTHR advocates, particularly those in Defence’s central policy community, to protect the project as a technological niche development despite inevitable technical problems and set backs. After achieving its limited objectives as an OTHR technology demonstrator, Jindalee Stage A was terminated in February 1979, some five years and a total expenditure of $6 million after its start in 1974. The cost incurred and
280 ibid., p. 24. 281 Joint Committee of Public Accounts, Report No 243 – Review of Defence Project Management, Vol 2 (Project Analyses), Austrlian Government Publishing Service, Canberra, 1986, p. 183. 282 Sinnott, The Development of OTHR in Australia, p. 23. 161
time taken to complete the Jindalee Stage A phase of OTHR development in Australia are summarised in Table 8.3.
Table 8.3: Jindalee Stage A
Dates Activity Time Cost incurred ($Am) Elapsed (years)
Apr 1974 – Jindalee Stage A 5 6
Feb 1979 (concept demonstrator)
Jindalee Stage B was intended to demonstrate the operational capabilities of OTHR in the Australian context. DSTO had begun planning for Stage B in 1974, at about the same time as the Minister approved Stage A. The planning entailed trade-off studies aimed at, for example, determining what combination of single dwell of the beam and repeated dwellings of the beam optimised the probability of detecting a target in a given state of the ionosphere. To conduct such studies the DSTO scientists required much more capable signal processing algorithms. This development was undertaken by DSTO in-house.
In May 1978 the Australian government approved a six-year Jindalee Stage B at an estimated project cost of $24.6 million. As this budget and schedule indicate, Stage B entailed a very substantial upgrading of the Jindalee Stage A radar. Key enhancements included the provision of a scanning beam covering a 60-degree sector and a much larger receiving array to enhance the system’s sensitivity. Processing the much larger volume of data generated, the scanning beam required much more sophisticated computing equipment. Other enhancements included much more sophisticated displays better suited to use by Service personnel and greater automation.283
Although the Stage B budget was some four times larger than that for Stage A, Strath still needed US support to achieve progress sufficiently quickly to ward off Service scepticism and retain the support of defence planners. In working up Stage B, Strath and his DSTO colleagues tested the progressive output of the Trade Off studies with US experts in April 1975, February 1976 and May 1977. Costs were also reduced by, for example, using second- hand HF power amplifiers provided by the US and re-using the Stage A transmitting arrays.
283 ibid., p. 32. 162
Finally, Stage B was also able to take advantage of leading-edge computing technology being developed by US companies.284
Access to US commercial technology was necessary to meet Stage B objectives but not sufficient to meet Stage B real-time signal processing requirements. In order to keep to the Stage B schedule, Strath decided to develop in-house a signal processor customised for OTHR. This required DSTO to develop a suitable computing language and software operating system as well as the processor digital hardware. All this took longer and cost more than anticipated.285 The processor so developed was ahead of what was then available commercially. Without the processor, the Jindalee Stage B system could not have achieved the level of performance required to demonstrate the military utility of OTHR in the requisite timeframe. But in designing and developing the processor in-house, DSTO again denied local companies the opportunity or incentive to develop the competencies that would be needed if and when the government decided to procure the system.
The impact of these and other technical difficulties on the Stage B schedule were exacerbated by the immaturity of the contemporary Australian defence sectoral innovation system. For example, software development was delayed for six months by DSTO’s inability to recruit the requisite computing professionals within Australian public service constraints. Similarly, obtaining delegate approvals for relatively low-value contracts for development, operation and maintenance for equipment and infrastructure took nine months.286 As a result, defence decision-makers agreed to Strath’s recommendation to slip the program by 15 months.
In order to evaluate the military utility of OTHR in the Australian context, Jindalee Stage B provided for a series of Jindalee Service Evaluation Trials (JSET). The trials were intended to test the radar’s detection and tracking performance in various tactical scenarios and in varying ionospheric conditions.287 To this end the trials ran intermittently from October 1984 to June 1986, providing a comprehensive demonstration of the military utility of OTHR for broad area surveillance of the nation’s northern maritime approaches, albeit at extra cost and schedule slippage – the program took nine years, rather than the six originally approved – and cost some $10.1 million more than the $24.6 million originally budgeted. Stage B operations continued until December 1987 at an overall cost of $34.7 million.
The cost incurred and time taken to complete the Jindalee Stage B phase of OTHR development in Australia are summarised in Table 8.4.
284 S. Colegrove, Project Jindalee: From Bare Bones to Operational OTHR, Proceedings of the Institute of Electrical and Electronic Engineers International Radar Conference, 2000, p. 823. 285 See Sinnott, The Development of OTHR in Australia, pp. 31-32 and Colegrove, Project Jindalee, p. 823. 286 Joint Parliamentary Committee of Public Accounts, Report No 243 p. 186 para 10.15. 287 Colegrove, Project Jindalee, p. 825. 163
Table 8.4: Jindalee Stage B
Dates Activity Time Cost incurred ($Am) Elapsed (years)
May 1978 – Jindalee Stage B 9 34.7
Dec 1987 (operational evaluation)
The Jindalee Stage B activity approved in 1978 also provided for eventual RAAF operation of the capability, preparation for which was managed by the Chief of Air Force Operations (CAFOP)288 who, at the time, was key advocate of the procurement of AEW&C aircraft for broad area surveillance. These organisational arrangements meant that, despite the favourable JSET results, the RAAF continued to contest OTHR’s capability value relative to AEW&C aircraft. In contrast, capability planners became increasingly convinced that OTHR proffered better value for money in Australia’s circumstances. By 1982 the RAAF was planning a limited operational conversion of the Jindalee Stage B radar aimed at obtaining an immediate OTHR capability and confining further investment in the Stage B radar to that required to address operationally deficient features of the experimental equipment.289 The planning did not address design constraints inherent in an experimental system. This minimalist option was formalised in Joint Staff Requirement (JSR) 13: A Requirement for an Over-the-Horizon Surveillance Radar and Data Dissemination System, issued in August 1983.
In October 1983 project definition studies were commissioned from AWA and CSA with a total budget of $0.6 million. The studies were intended not only to define the conversion in more detail but also to prepare the way for greater Australian industry involvement in OTHR technology.290 In the event, the studies were pre-empted by the JSET after the expenditure of some $0.3 million. Thereafter, they seem to have made little impact on OTHR outcomes.
Beginning in 1982, the RAAF began posting RAAF pilots and engineers to the project to prepare the way for operational conversion by working on Jindalee Stage B. DSTO also transferred control of the Jindalee Stage B assets to the RAAF which renamed it JFAS. As these preparations for a minimalist conversion of the Jindalee Stage B radar for operations progressed, senior Defence officials took steps to change the trajectory of the OTHR innovation. Preparations for a minimalist conversion of the Jindalee Stage B radar were eventually overtaken by the announcement in October 1986 by the Minister for Defence of the government’s decision to procure a network of OTHRs, to be called the JORN.
288 Joint Parliamentary Committee of Public Accounts, Report No 243, p. 185, para 10.11. 289 Sinnott, The Development of OTHR in Australia, pp. 37-38. 290 Sinnott, The Development of OTHR in Australia, p. 38. 164
The Minister’s announcement led to a redirection and reinvigoration of investment in JFAS. The Facility continued to be operated by both DSTO and the RAAF as a test bed for ongoing development of both scientific and operational aspects of OTHR. This development continued for decades, sustained by a new generation of RAAF leaders who embraced OTHR as part of a network-enabled air defence capability and by DSTO commitment to improving the military utility of OTHR. Separately, and in parallel with the JFAS activity by DSTO and the RAAF, the DAO began the process of procuring JORN as foreshadowed by the Minister.
The cost incurred and time spent on the minimalist OTHR option based on JFAS are summarised in Table 8.5.
Table 8.5: Jindalee Facility Alice Springs
Dates Activity Time elapsed Cost incurred (years) ($Am)
Aug 1983 – Oct JFAS 3 0.3 1986 (minimalist option)
8.4 Formulating demand for an OTHR-based broad area surveillance capability This section traces the process by which the customer element of the Australian radar defence competence bloc formulated the demand for an OTHR-based solution to Australia’s requirement for a cost-effective broad area surveillance capability. The process began with growing concern among senior Australian defence actors at the perfunctory approach to exploiting the military utility of OTHR. This concern and the need to ensure Air Force cooperation resulted in the civilian Secretary and the military Chief of the Defence Force jointly commissioning a report on OTHR options. The preliminary findings of this report were then endorsed as part of a separate and much wider review of Australian defence capabilities. The process culminated in the announcement by the Minister for Defence of the government’s decision to procure the JORN in October 1986. Because the above process illustrates how the Australian defence innovation system worked at the time, it is explained in more detail in the following paragraphs.
165
In October 1985 the Secretary of the Department of Defence and the Chief of the Defence Force commissioned the Acting Head of the predominantly civilian Force Development and Analysis Division, Dr Mike Gilligan, to prepare a report on options for Jindalee/OTHR.291 Gilligan, with the support of a small team of officials from Air Force, DSTO, Navy and the Capital Procurement Organisation, was to canvas alternatives and provide costed options for the future development of OTHR in accordance with national strategic interests.292
In the course of preparing their report, Gilligan and his team visited the UK and the US. In the UK, they spoke to GEC Marconi, Britain’s leading radar supplier who was to figure prominently in JORN procurement but who, according to Gilligan, had no OTHR expertise.293 In the US Gilligan and his team talked to the GE/TRW team who were building the AN/FPS- 118 radar for the USAF and who, in Gilligan’s view, had by far the most technical and managerial expertise to offer Australia. Because a recommendation to this effect was outside his terms of reference, and because source selection was a matter for the DAO, Gilligan stopped short of actually recommending that Australia ask GE/TRW to lead the development of JORN.
Gilligan and his team had completed their highly classified report in mid-1986. The Gilligan report acknowledged that OTHR should not be seen as the sole sensing element of an air defence system, recognised that in due course it would need to be augmented by other assets but then rebutted the prevailing minimalist logic. It did so by providing a rigorous and detailed operational analysis of the potential military utility of a network of OTHR in various defence of Australia contingencies. The report concluded that OTHR provided a degree of transparency of Australia’s ocean and air approaches that added new dimensions to Australia’s defences. It also pointed out that, subject to development of satisfactory command and control arrangements, OTHR stood to enhance the tactical utility of Australia’s air and surface defence assets including, notably, the F/A-18.294
Gilligan then played the role of capability entrepreneur in recommending that priority be given to developing a network of radars coordinated around the continent to provide high assurance and early warning in key strategic areas. He further recommended that DSTO continue a substantial long-term R&D program into OTHR, but that DSTO’s initial priority should be working with industry to design a new radar, ensuring that sufficient technology was transferred to enable independent indigenous production of the OTHR and its through-
291 P.H. Bennett and R.W. Cole, Study of Options for Jindalee/OTHR, Memorandum CDF 902/1985 – SEC685.1985 dated 20 October 1985 (Bennett) and 28 October 1985 (Cole) in M.F. Gilligan: A Report to Secretary and CDF on Over-the-horizon Radar. 292 Terms of Reference for a Study of Options for Over-the-Horizon Radar and Project Jindalee, attached to Bennett and Cole Memorandum CDF 902/1985 – SEC685.1985, available from [email protected]. 293 M.F. Gilligan, Transcript of Hearings by Joint Committee of Public Accounts and Audit into the Jindalee Operational Radar Network, 5 December 1996, p. 25. 294 M.F. Gilligan, A Report to Secretary and CDF on Over-the-horizon Radar, Executive Summary, unpublished, p. 2, para 6. 166
life support. Finally he recommended that Defence conclude a broad agreement on long- term OTHR information exchange with the US, covering operational, trials and technology aspects.295 In recommending that Australia acquire a network of OTHR, Gilligan emphasised that the viability of that recommendation rested on Australian access to US technical and managerial expertise.296 In the event, this fundamental element of Gilligan’s case for JORN was ignored.
The Gilligan report was, necessarily, highly classified. While such highly classified analysis was necessary for internal defence debate, it was of limited usefulness in building public support for investment in such a capability. Accordingly, in early 1986 Gilligan again played capability entrepreneur. He briefed Professor Paul Dibb, then in the final stage of preparing his seminal review of Australian defence capabilities. On the basis of this briefing, Dibb included the following statement in his report:
This Review judges that three radars could be justified in locations across the continent. This is a priority matter and detailed attention should be given to this development with the objective of a decision in FY 1987-88 to allow two additional radars to enter service by the early 1990s. The design of these radars should allow for later modification to take advantage of technological development as it occurs, but these possibilities should not be used to delay their early introduction into service.297
The Dibb review of Australia’s defence capabilities, released in March 1986, prepared the way for the Minister’s JORN announcement in October 1986. During the intervening period, however, the debilitating intra-defence disputation about the capability value of OTHR became public. In June 1986, for example, Australia’s national newspaper published an article to the effect that the Gilligan report was based on unsubstantiated claims of Jindalee’s performance, that it was almost certain to cause the military to revolt over what they saw as an increasing civilian encroachment into operational areas and that Jindalee’s exaggerated capabilities were being used to justify further deferral of the purchase of AEW&C aircraft.298 In July 1986 the recently retired Chief of Air Force echoed these views in a widely publicised speech.299
The intra-defence disputation continued despite the announcement in October 1986 by the Minister for Defence of the government’s decision to acquire a network of OTHRs. In March 1987, for example, the press continued to report ongoing internal dispute about the relative priority to be accorded to AEW&C aircraft and OTHR in the competition for funds in the
295 ibid., pp. 9-12. 296 ibid., p. 27. 297 Dibb, Review of Australia’s Defence Capabilities, p. 117. 298 Peter Young, A hard look at Jindalee, Australian, 23 June 1986, p. 2. 299 David Evans, Air Power in the Defence of Australia: the Strategic Context, in Des Ball (ed.) Air Power: Global Developments and Australian Perspectives, Pergamon Press, Rushcutters Bay, 1988, p. 124. 167
absence of a threat to Australia.300Overall, intra-defence disputation probably added some two years to the JORN development and acquisition schedule. More generally, however, Australian companies had little incentive to invest in OTHR-related competencies in the face of this kind of disputation within the customer element of the Australian defence competence bloc.
In October 1986 the then Minister for Defence, Kim Beazley announced that the government had decided to proceed with the design and development of a network of OTHR. As promulgated in the 1987 Defence White Paper, the formal strategic rationale for the decision emphasised the value accorded effective surveillance of Australia’s sea and air approaches and the prohibitive cost of relying on AEW&C aircraft alone to provide such surveillance. It concluded with the judgement that, “For Australia, OTHR with its ability to sweep large volumes of air and sea space from a single location offers the only affordable solution.”301
The time taken by the Australian defence customer in changing from JFAS to JORN are summarised in Table 8.6. The cost incurred in doing so is subsumed in the costs for JFAS upgrades summarised in Table 8.8.
Table 8.6: From JFAS to JORN
Dates Activity Time elapsed (years) Cost incurred ($Am)
Oct 1985 – Oct 1986 Gilligan review, Dibb 1 Not applicable302 review
8.5 Executing demand for an OTHR-based broad area surveillance capability The procurement of an OTHR-based solution to Australia’s requirement for a broad area surveillance capability did not conform to the conventional pattern of Australian defence capital equipment procurement. Conventionally, the Australian defence customer approached the market with a requirement, candidate suppliers proffered competing solutions to the requirement and the customer selected that solution it judged offered best value for money. In JORN’s case, however, DSTO innovators first developed the OTHR technology, capability planners then secured agreement to use it for broad area
300 Patrick Walters, RAAF in war with Navy on spending, Sydney Morning Herald, 11 March 1987, p. 7. 301 Beazley, Kim: The Defence of Australia 1987, Australian Government Publishing Service, Canberra, 1987, pp. 34-35. 302 See Table 8.8: JORN post-acceptance integration and upgrade in this chapter. 168
surveillance, leaving the DAO to identify and engage an industrialist to produce an OTHR- based solution.
Accordingly, this section analyses how the DAO executed demand for an OTHR-based solution to the Australian requirement for a broad area surveillance capability. The section begins with a description of the process by which the DAO searched for an industrialist able to design, develop and produce the JORN envisaged by the government and as announced by the Minister for Defence. The section then describes the process by which the DAO selected, from the candidates so identified, that industrialist whose bid for designing, developing and producing JORN was judged, ex ante, to offer best value for money. The section concludes with a discussion of the process by which the DAO procured, from the industrialist so selected, a JORN that not only met the requirement but did so in a way that realised, ex post, the value for money anticipated by the government ex ante.
The search by the DAO for a JORN industrialist was conditioned by, firstly, the performance parameters set by the capability planners in the JORN Operational Performance Directive. The search was also conditioned by, secondly, Australian defence policies regarding Australian Industry Involvement (AII) in defence capital equipment procurement and by, thirdly, Australian defence policies regarding Australian Ownership, Control and Influence (AOCI). The DAO conducted the search in accordance with the prequalified tender process.
The request for tender specified a network of OTH radars able to meet significantly more demanding requirements than the JFAS radar. These enhanced requirements were stipulated in the JORN Operational Performance Directive. They included, for example, greatly expanded radar coverage combined with more accurate target location; equipment for transmitting, receiving and processing signals able to operate with much reduced electrical noise; integration of the two radars into a single network and greater resistance to electronic interference (including attempts by an adversary to jam the system). The OTHR technology developed under DSTO auspices in the early phases of Jindalee required substantial development to meet these new requirements. Hence, in pursuing a local ‘make’ solution to JORN, local companies were not only expected to absorb DSTO technology, they were also required to embellish it substantially.
The DAO’s search for, and selection of, a JORN prime contractor was conditioned by the Australian government’s policy for AII in defence procurement. During the 1980s, AII was geared to defence self-reliance by inclusion of provision for Defence Designated and Assisted Work (DDAW). This provided for elements of the item being procured to be designated for manufacture, assembly, test and set-to-work in Australia. The JORN project included provision for both AII and DDAW. In briefing industry, the DAO acknowledged that JORN-related DDAW may well entail cost and delivery time penalties. Companies had to justify such penalties on a case-by-case basis and in terms of their contribution to self- reliant defence capability. DDAW was complemented by the Defence Offsets Policy. This 169
policy required overseas suppliers of goods and services worth $2.5 million or more to place in Australian industry technology transfer or work to the value of 30% of the imported content of those goods and services.303
In the mid-1980s OTHR was one of several sensitive indigenous technologies then being developed in Australia that the Australian defence customer wished to hold particularly closely. To provide such protection, the Australian defence customer invoked a new policy called AOCI. This policy restricted access to such specially designated technology to Australian nationals and to enterprises that could demonstrate a very high level of Australian control of their local operations.304 Such extra restrictions were over and above companies’ extant obligations to protect Australian government-classified information and to comply with stringent US requirements regarding third party access to US technology released to Australia. The AOCI policy was also a consequence of Australia’s earlier difficulty in obtaining US government agreement to release the software source codes Australia considered necessary for self-reliant operation of the F/A-18 aircraft.305 The AOCI policy effectively annulled any prospect of a GE/TRW-led development of JORN as Gilligan had envisaged. The AOCI policy also caused the DAO to use a combination of prequalified and limited tendering procedures in approaching the market.
The DAO began its search for a prime contractor in May 1988 by inviting Australian companies to register interest in tendering for the JORN contract. Companies registering interest were required to hold Australian government security clearances sufficient to permit access to classified JORN tender documentation. In addition, because the JORN project fell within the ambit of the AOCI policy, potential contractors were required to comply with the associated management, staffing and ownership arrangements. Taken together, JORN security constraints and the AOCI policy meant that, effectively, the DAO would only consider awarding the JORN prime contract to Australian-owned and Australian- controlled companies.
In imposing these conditions the DAO was well aware that Australian companies had little, if any, competence in OTHR and, indeed, limited competence in undertaking a development project of JORN’s scale and complexity. Accordingly, the DAO intended that candidate Australian prime contractors would team with overseas partners within the constraints of the AOCI policy.306 This was conveyed to some 50 Australian companies who responded to the DAO’s invitation to register interest in JORN and who attended associated industry briefings conducted by the DAO in 1988. The head of the DAO (called Deputy Secretary Acquisition – DEPSEC A) was delegated authority by the Minister for Defence to restrict
303 Beazley, The Defence of Australia 1987, p. 80. 304 ibid., pp. 82-83. 305 See also testimony by M. Gilligan, in Hansard record of hearings by the Joint Committee of Public Accounts (JCPA), Review of the JORN Project, Canberra, 5 December 1996, p. 29. 306 Testimony by M. Brennan, in Hansard record of hearings by the JCPA, Review of the JORN Project, Canberra, 6 December 1996, p. 44. 170
defence business to certain companies on a case-by-case basis. In 1988 the then DEPSEC A, Dr Malcolm McIntosh, exercised his delegation to confine participation in the project definition studies to three companies shortlisted from the 50-odd who had registered interest in JORN. The three companies so shortlisted were Telecom (later renamed Telstra), the national telecommunications carrier who assigned the task to their Applied Technologies Division; AWA (who, as discussed earlier, had previous OTHR involvement); and BHP Aerospace and Electronics Ltd (a wholly owned subsidiary of what was then Australia’s largest company, Broken Hill Pty Ltd (BHP), whose core business was steel, mining and petroleum).
The Project Definition Studies undertaken by these companies were funded by Defence, begun in September 1988 and completed in March 1989.307 On 15 May 1989, following completion of the above Project Definition Studies, Dr McIntosh again exercised his delegation on behalf of the Minister for Defence to restrict the JORN request for tender to Telstra, AWA and BHP.308 These three companies were required to submit tenders by late August 1989.309 Each tenderer was required to nominate a team of local and overseas sub- contractors and each was required to submit target price and price ceiling incentive bids.310
As already indicated, the DAO recognised that the Australian prime contractor for JORN would rely on technology input from an overseas sub-contractor. The DAO envisaged that overseas sub-contractor transferring its technology to the Australian prime contractor in such a way that the latter would not only become knowledgeable and competent in the technology but that it would absorb the technology sufficiently to be able to undertake future defence contracts of comparable complexity.311 The three tenderers shortlisted on this basis comprised AWA, BHP and Telecom.
AWA had a long history of involvement in supply and support of Australian defence electronics, including the Jindalee experimental radar since 1984. AWA teamed with General Electric of the US. General Electric had designed, developed, constructed and set to work the US Air Force’s AN/FPS-118 OTHR radar. The AWA team also included CSA (who had also developed and supported software for Australian defence projects, including JORN) and Transfield (who had a proven record in managing large Australian defence projects, including the ANZAC frigates).
BHP was Australia’s biggest company, with interests in mining and steelmaking. It had established an Aerospace and Electronics Division which had secured the licence to build
307 Department of Defence, JORN Project Management and Acquisition Plan (Section 2), April 1993, p 2-9 of extract tabled by the JCPA (see JCPA Exhibit 9). 308 ibid., p. 2-10 309 K. Beazley, Tenders Invited for Over-the-Horizon Radar Network, Ministerial Press Release no 113/89 of Thursday 1 June 1989. 310 McNally, Jindalee Operational Radar Network, p. 42, para 6.15. 311 Testimony by Brennan, p. 44. 171
and operate the laser airborne depth sounder developed by DSTO. BHP teamed with the Raytheon Company who had built the US Navy’s AN/TPS-71 ROTHR system. BHP’s team also included certain Australian software development and engineering companies.
Telecom was an Australian government-owned corporation descended from the Commonwealth Government’s Post Master General (PMG) Department. It retained the PMG’s monopoly on provision of telecommunications in metropolitan and regional Australia. At the time of the JORN tendering process, Telecom’s exposure to Defence work was essentially confined to provision of secure video conferencing the ADF’s major joint force exercises. In tendering for the JORN contract, Telecom teamed with a British consortium GEC-Marconi. The latter had expertise in radar systems but not in OTHR. However, GEC-Marconi’s Australian subsidiary had supplied and installed high-powered, HF transmitters (designed and built in the UK) at the JFAS. The Australian government later privatised Telecom which was renamed ‘Telstra’. To avoid confusion in the discussion below, this thesis has used the name ‘Telstra’.
The DAO did not attempt to influence which Australian company teamed with which overseas company and on what basis. In the prevailing defence project governance, such teaming arrangements were considered a commercial matter which, in accordance with the norm of open and effective competition, was beyond the scope of proper DAO influence. The DAO required tenderers to bid for JORN on a turnkey basis. This entailed Defence specifying the JORN functionality it wanted and then looking to the prime contractor to engineer, design, develop, construct, install, set-to-work, test, document and offer for acceptance a JORN conforming in every respect to those specifications.312 An important discriminator among competing bids was the extent to which each tenderer was prepared to assume the cost, schedule and technical risk inherent in a turnkey project.
Representatives of the DAO, RAAF and DSTO began examining the three tenders in September 1989 and in early 1990 eliminated the BHP bid. Later that year the DAO invited the AWA/GE and Telstra/GEC-Marconi consortia to submit new tenders addressing a revised requirement for two radars (one near Laverton in Western Australia and one near Longreach in Queensland), as well as a network/coordination centre at RAAF Base Edinburgh in South Australia.313
During this second round of tendering activity, the US press reported the DAO was favouring the Telstra/GEC-Marconi bid because the US government was reluctant to guarantee, ex ante, release of OTHR technology.314 The US press report prompted an exchange of correspondence between US and Australian defence officials. In the course of this
312 N. Hammond, Review of Auditor General’s Report 28, letter FASDM 456/96 of 27 November 1996, tabled as Exhibit No 17 in the JCPAA Review of the JORN Project. 313 Department of Defence, JORN Project Management and Acquisition Plan, p. 2-10. 314 See Defense News, 5 February 1990 and Canberra Times, US ‘delays’ Aust radar project, 14 February 1990, p. 16. 172
correspondence Dr McIntosh acknowledged US assistance with OTHR-related R&D but noted the importance to Australia of having in country the ability to build, operate, maintain and adapt the system to meet current and future requirements. To that end, McIntosh emphasised, Australia needed to have full control over the system, software design and codes, and other aspects that affect its operation.315
On 20 December 1990, the then Minister for Defence, Robert Ray, announced that Telstra (Telecom) had been chosen as the preferred tenderer for the construction of the above system.316 In judging that the Telstra bid offered better value for money, the relevant Australian defence actors were aware that both Telstra and GEC-Marconi, its key sub- contractor, knew less about OTHR than the rival team comprising AWA and General Electric.
In assessing the value for money of the AWA/GE bid, the Australian defence customer probably took into account the above concern over the US government’s willingness to allow General Electric to transfer the technology that AWA required to produce an operational JORN system that conformed with the JORN Operational Performance Directive.317 It is unlikely, however, that such concerns were the only, or even the most important, consideration in judging the relative value for money of the AWA/GE and Telstra/GEC-Marconi bids. Three other considerations also influenced judgements by the Australian defence customer as to the two bids’ relative value for money. Firstly, the Australian defence customer would have taken into account the greater willingness of the Telstra/GEC-Marconi team to accept more project risk than its rival. Secondly, the Australian defence customer accorded high value to the digital technology proposed by Telstra/GEC- Marconi team which offered greater long-term growth potential. Thirdly, AWA was then experiencing certain commercial difficulties that, in the opinion of key actors in the customer element of the Australian defence competence bloc, reflected adversely on that company’s ability to assume prime contractor responsibilities. As the way these actors interpreted AWA’s commercial difficulties sheds useful light on the workings of the Australian defence procurement process at the time, AWA’s difficulties are explained in more detail below.
In 1987 AWA reported that it had sustained $49 million worth of foreign exchange losses due to unauthorised trading by an AWA employee in 1986 and 1987. AWA subsequently initiated legal proceedings against auditors for failing to identify the trading. The auditors, in turn, initiated cross claims against the company's directors. AWA also took out an injunction
315 M. McIntosh, unpublished correspondence, reference DEPSEC A&L 33/1990 of 8 February 1990. 316 Robert Ray, Government approves construction of Jindalee Over the Horizon Radar Network, Ministerial Press Release no 201/90 of 20 December 1990. 317 This judgement is corroborated by comments by the then Secretary of Defence – see testimony by T. Ayers, in Hansard record of hearings to the JCPA Review of the JORN Project, Canberra, 6 December 1996, p. 83. 173
against the foreign exchange trader and the banks involved.318 These developments were in play by 1987-90, just as the DAO was soliciting and evaluating JORN bids.
The DAO gave considerable weight to AWA’s commercial difficulties in judging, ex ante, the ability of the AWA company in general, and the ability of AWA’s senior managers in particular, to assume the kind of prime contractor responsibilities envisaged under AOCI. In judging (ex ante) the relative efficiency of the AWA and Telstra bids, the DAO officials involved would have also been cognisant of the fact that ex-post access to DSTO’s expertise by JORN industrialists was a fundamental tenet of the JORN contract. This suggests that those officials envisaged Telstra’s access to DSTO expertise helping to compensate for the paucity of Telstra/GEC-Marconi’s knowledge of OTHR technology relative to that of AWA/General Electric, thereby reducing the weight they accorded to AWA’s greater technical competence in their assessment of the relative merits of the two bids.
The procurement of JORN took far longer and cost much more than originally anticipated, in the course of which Telstra/GEC-Marconi relinquished the contract to another company, RLM. To ascertain what this indicates about the Australian defence innovation system, the following discussion of the JORN procurement process is divided into an initial phase, marked by Telstra’s stewardship as prime contractor, a transition phase, marked by Telstra relinquishing the prime contract and its takeover by RLM and a completion phase, marked by RLM’s stewardship as prime contractor.
The initial JORN procurement phase began with signing of the initial JORN procurement contract by the DAO and Telstra on 11 June 1991. The JORN initial procurement phase was characterised by technical difficulties, schedule slippage and cost blowouts.319 While these problems can be attributed in part to the lack of technical competence on the part of Telstra and GEC-Marconi, they were exacerbated by the customer’s interpretation of the turnkey contract, by the allocation of risk between customer and prime contractor and by the prime contractor’s unwillingness and inability to access DSTO’s expertise in OTHR. Under the turnkey contract concluded between the DAO and Telstra, Defence defined what Telstra was to deliver in terms of functions to be performed and standards to be achieved in the performance of those functions – see discussion of JORN Operational Performance Directive in sub-section 10.4.1 above. Telstra, on the other hand, undertook to “engineer, design develop, construct, install, set to work, test, document and offer for acceptance the Jindalee Operational Network conforming in every respect with the specifications”.320
In focusing on functional specifications, Defence deliberately eschewed issuing detailed requirements specifications. That step was a matter for Telstra as prime contractor:
318 Trevor Sykes, Anatomy of a $50million loss, Bulletin, 19 January 1993, pp. 76-73. 319 As documented in McNally, Jindalee Operational Radar Network. 320 ibid. 174
it was a tenet of the contract that the risk for the contractual outcome resided with the contractor, and therefore Defence went to some lengths to make sure that its involvement in the contract and its actions in the contract did not transfer risk from the contractor to the Commonwealth.321
This approach to the turnkey contract led Defence’s JORN Project Office (JPO) to decline to provide progressive formal approval of the work undertaken by the prime contractor. The DAO/JPO logic was that if Defence had given approval for, say, a particular design or a particular process which, when it was subsequently implemented, proved to be unacceptable, then the contractor could have claimed that it was acting in accordance with Defence approval and, hence, had no liability.322This same logic worked to inhibit Telstra’s willingness and ability to take advantage of DSTO expertise as DAO originally envisaged in setting up the JORN procurement.
Unable to rectify their known deficiencies in OTHR technology by leveraging DSTO expertise, Telstra and GEC-Marconi had to build the requisite competencies. This proved to be an insurmountable task. Telstra had to build, from scratch, the business acumen required to deal with, on one hand the Defence customer and, on the other hand, GEC-Marconi as key sub-contractor. Telstra also needed to grow the technical knowledge required to relate the JORN Operational Performance Directive to GEC-Marconi proposals and to develop the project management competencies to manage a complex development project. GEC- Marconi, which had no prior experience in OTHR, had to develop the technical knowledge required to develop an OTHR solution that complied with the JORN Operational Performance Directive.
The time Telstra and GEC-Marconi took to build those competencies, and the cost they incurred in doing so, completely overwhelmed the structure of incentives flowing from the price ceiling incentive contract the DAO negotiated with Telstra. This contract provided for a target price of $685.5 million (April 1991 prices). It also provided for a maximum (ceiling) price payable by the Commonwealth equal to the target price plus 60% of any cost overruns up to a maximum of 10% above the target price (that is, a ceiling price of $741.1 million in April 1991 prices). Telstra’s incentive to achieve the target price and to avoid breaching the ceiling price was reinforced by arrangements to share the downside financial risk and the upside savings risk. Under the financial risk-sharing arrangement, Telstra was responsible for 40% of any cost overruns up to the ceiling price, and 100% of all costs that exceeded the ceiling price. Under the savings-sharing arrangement Telstra was entitled to 40% of the savings if JORN was completed for less than the target price.323
321 Testimony by Brennan, p. 49. 322 ibid. 323 McNally, Jindalee Operational Radar Network, p. 42 para 6.15. 175
Under the above price ceiling incentive contract, Telstra assumed the risk inherent in cost overruns while Defence bore the risk of delay, manifest in terms of capability foregone. According to a senior Defence official, Defence recognised this schedule risk from the outset of the project:
we did recognise that there was significant risk in this project and in many ways we felt one of the ways the risk would manifest itself would be in schedule. So we constructed the framework of incentives and disincentives in a way that recognised that. There is a very real cost to the contractor in running late in terms of extra overheads. There are, however, real costs to the Commonwealth in the capability foregone.324
Deficiencies in both Telstra and GEC-Marconi competencies and DAO business mistakes meant that both cost and schedule risk materialised. Telstra/GEC-Marconi proved unable to develop the competencies required to produce a compliant design at a price and in a timeframe specified in the turnkey contract. In June 1996 the ANAO observed that:
With 80 per cent of the JORN prime contract target price spent (or 73 per cent of the ceiling prince) and 80 per cent of the schedule elapsed, less than 20 per cent of the configuration items have passed the critical design review stage. Current plans indicate that delivery of JORN, which the contract still specifies as June 1997, will not occur until June 2000. If current payment trends continue Defence’s JORN full-scale development budget will be spent by mid-1997 (the original project completion date), but there will still be two or more years of system development work to be done including the high risk high cost systems integration phase. It is therefore apparent that the JORN project will surpass the target price and reach the ceiling price. The contract provides that, after that latter point, additional costs will be borne by Telstra.325
Telstra (and its sub-contractors) were expected to draw on the knowledge DSTO accumulated in establishing and operating Jindalee Stage A and Stage B. Such DSTO information and expertise was formally identified in the JORN prime contract as background IP which, immediately upon creation, vested in and became the property of, the Commonwealth. In order to comply with the JORN Operational Performance Directive, however, Telstra was obliged to generate JORN improvements, defined as JORN-related IP rights, created by a contractor and funded by other than the Commonwealth. Under the terms of the JORN prime contract, all IP, including JORN improvements, became the property of the Commonwealth. In return for a $2.4 million reduction in Telstra’s tender, however, the Australian defence customer granted Telstra an irrevocable, world-wide,
324 Testimony by G. Jones, in Hansard record of hearings to the JCPA Review of the JORN Project, Canberra, 23 July 1996, p. 19. 325 McNally, Jindalee Operational Radar Network, p. 21 para 3.20. 176
royalty-free and unrestricted licence to use, commercially exploit or sub-licence JORN IP, for 20 years, unless such action was contrary to the national interest of the Commonwealth.326
The above national interest provision had far-reaching implications for the pattern of JORN development after it had been accepted into RAAF service. From the outset of the JORN project, Defence saw little scope for Telstra (or its successors) to export complete JORN systems because of the operational sensitivity of some JORN systems and sub-systems.327 The subsequent Parliamentary hearings reveal acute concern by both parliamentarians and officials that DSTO IP might have diffused via Telstra to GEC-Marconi, pre-empting any opportunity for local companies to gain commercial benefit.328 The research undertaken for this thesis found nothing that would indicate how Telstra envisaged using the above licence to commercialise JORN IP. It is reasonable to conclude that the pattern of JORN development continued to be driven by Australian defence demand.
As Telstra began to encounter difficulties with the project, its actual losses began to accumulate and its prospective losses to mount. This caused Telstra and GEC-Marconi to initiate a JORN Technical Audit by experts from what was then Lockheed Martin’s Naval and Electronic Sensor Systems Division. The audit, which was completed in September 1995, was a damning indictment of JORN project management. Telstra eventually passed a copy to the JORN Project Office in November 1995.
In accordance with Australian defence project governance, the ANAO had full access to the Lockheed Martin audit as well as Defence records in auditing the project in 1995-96. As already indicated, the ANAO published a scathing performance audit of Defence’s management of the JORN project in June 1996. The ANAO report prompted the Parliament’s Joint Committee of Public Accounts and Audit (JCPAA) to hold a public hearing to examine Defence management of the project in July 1996. The JCPAA was dissatisfied with information provided by Defence and other witnesses at that hearing and decided to hold its own enquiry. To this end, the JCPAA invited submissions and held hearings intermittently over July 1996 to March 1997.
All this activity and the associated press coverage alerted commercial entrepreneurs to a prospective business opportunity. One such entrepreneur was Paul Johnson, at the time the Managing Director of Lockheed Martin’s Asia Pacific business.329 In conducting the above technical audit, Lockheed Martin had gained detailed insight into Telstra/GEC-Marconi’s handling of the JORN project. Importantly, however, Lockheed Martin’s Naval and Electronic Sensor Systems Division had previously been owned by General Electric and had built the USAF’s AN/FPS-118 OTH radar. Before General Electric sold the Division to Lockheed Martin,
326 Joint Committee of Public Accounts and Audit, Report No 357, pp. 106-107. 327 Beazley, The Defence of Australia 1987, p. 85, para 6.62. 328 Joint Committee of Public Accounts and Audit, Report No 357, pp. 104-116. 329 The following information is based on Paul Johnson (CEO Lockheed Martin Australia), interview, Canberra, 25 July 2005. 177
the Division had provided the technical core of the AWA/General Electric team that bid unsuccessfully for JORN in 1989.
Johnson had participated in the AWA/General Electric bid in his capacity as Marketing Manager Australasia of General Electric Aerospace. He had stayed in touch with JORN project developments before and after being appointed Managing Director of Lockheed Martin’s Asia Pacific business. In 1995 Mr Johnson secured Lockheed Martin’s agreement to ascertain whether Defence (as JORN buyer and operator) and Telstra (as JORN prime contractor) would be receptive to an unsolicited proposal that, subject to satisfactory due diligence, Lockheed Martin take over management of JORN from Telstra. In formulating this unsolicited proposal, Mr Johnson and his principals in Lockheed Martin obviously drew attention to the company’s proven scientific, engineering and managerial record in designing, developing, building and setting to work the USAF AN/FPS-118 OTHR system. In addition, however, they envisaged assigning to the JORN task scientists and engineers made available as a result of the USAF’s cessation of its OTH-B program in 1993.
Having obtained the support of Lockheed Martin principals for his initiative, Mr Johnson then sought Defence advice as to whether, given the previous policy position, the Australian government would accept an American company playing a pivotal role in the management of JORN. Defence advised Lockheed Martin to form an alliance with an Australian-owned, Australian-controlled partner in order to help ensure that key JORN technology and know- how remained in Australia. Lockheed Martin found a suitable partner in Tenix, by then a successful builder of frigates for the RAN but not a systems house. Tenix’s owners were prepared to share the JORN cost, schedule and technical risk in return for a commensurate share of the profits if JORN succeeded and with an eye to accessing the defence systems market.
In 1996, after gaining Defence and Telstra agreement in principle to taking over the contract, Lockheed Martin and Tenix formed a 50/50 joint venture. The latter was solely intended to provide the vehicle by which the two principals in the joint venture conducted the due diligence required to prepare a proposal to take over management of the contract from Telstra. In February 1997 Telstra concluded a contract with the joint venture to take over management of JORN for a fee of $64 million. Telstra entered into this arrangement as a first step in cutting its losses, relinquishing the JORN project and extricating itself from Australian defence business. This reflected a fundamental change in Telstra’s commercial strategy driven by growing pressure from private telecommunications suppliers and the government’s announced policy of privatising Telstra.
The due diligence process confirmed to Lockheed Martin and Tenix principals that JORN was robust in scientific and engineering terms. At issue was whether the joint venture would be able to reconfigure the JORN project on a commercially viable basis and realign the commercial interests of Telstra, GEC-Marconi and other contractors, having regard to the 178
major problems identified in project organisation and stakeholder relationships; project management; technical and management skills; absence of a requirements baseline; and shortfalls in software development, integration and test planning. The key problem was the flawed contracting arrangements between Telstra and GEC-Marconi. The contract motivated GEC-Marconi to deliver components and sub-systems as specified, irrespective of whether or not they worked effectively at the JORN system level. Because JORN was a developmental project, the gap between contract-specified deliverables and the emerging hardware and software requirements widened inexorably. The contract between Telstra and GEC-Marconi gave GEC-Marconi limited incentive to contribute effectively to the crucial system integration, test and set-to-work phase of JORN.
As a first step in determining what would be required to align GEC-Marconi’s interests with JORN project outcomes, the joint venture extended the 1996 Telstra due diligence to include GEC-Marconi. Fortunately for Johnson and Lockheed Martin, GEC-Marconi was then experiencing commercial difficulties that prompted the company’s owners to put the company on the market. What would otherwise have been a protracted due diligence exercise with an uncooperative GEC-Marconi was pre-empted by GEC-Marconi’s owners accepting an offer by British Aerospace. This gave GEC-Marconi a compelling incentive to liquidate its JORN liabilities.330 This cleared the way for the principals of the joint venture to begin a second round of due diligence with a view to taking over the project. In June 1997, as a first step to taking over the project, the joint venture was converted into a jointly owned company called RLM. This activity marked the end of the initial phase of the JORN procurement process and initiated the transition phase of that process.
RLM and its owners Lockheed Martin and Tenix were able to take advantage of commercial pressures on both Telstra and GEC-Marconi to obtain their agreement for RLM to assume full control of, and responsibility for, the project on satisfactory commercial terms. Defence then accepted the joint venture’s proposal that it take over management of JORN, re- baseline the project and then assume full contract responsibility for delivery of a remediated JORN project. This decision marked the beginning of the transition phase of the JORN procurement process. In agreeing to RLM’s taking over the contract, Defence satisfied itself that the new joint venture was in the defence business for the long term and that it had significant prospects for winning new defence business.331 Defence did require the parties to the new joint venture to make special arrangements for the protection of sensitive Australian classified information and intellectual property. However Defence stepped back from the stringent provisions of the earlier AOCI program which was discontinued in 1995 in favour of similar arrangements tailored to the specific circumstances
330 ibid., p. 6. 331 Hammond, N.: JORN project: (Defence) Submission to the Joint Committee of Public Accounts forwarded on 28 November 1996, available in JCPAA collected submissions, page S136, para 34. 179
of each case.332 By November 1997 RLM was operating to a re-baselined plan for the project and in November 1999 Defence novated the JORN contract to RLM. Telstra exited Australian defence business, having incurred a loss of $606 million.333 This marked the end of the transition phase of the JORN procurement process and initiated the completion phase of that process.
Defence took advantage of the novation process to secure RLM agreement to some key changes to JORN contractual arrangements. While progress payments were retained, Defence insisted on linking payments to earned value and dispensing with previous milestone payments system. In parallel, RLM instituted far-reaching changes to organisational arrangements, engineering processes and system integration processes. Of particular significance was RLM’s decision to invest $50 million of its $64 million Telstra fee to establish the Melbourne Integration Facility (MIF). RLM’s short-term intention was to facilitate the simultaneous development, integration and testing of JORN computing configuration items, sub-systems and systems. To this end the MIF could accommodate 350 integrated project team engineers in 20,000 square feet of integration and test laboratories, all fully instrumented and secure to render JORN’s evolving functions and performance fully transparent. Satellite communication links to South Australia and remote sites allowed RLM and members of the integrated project team access to and control of the remote radars during the integration and test phase.334
The MIF was the key to the plan by Lockheed Martin and Tenix to develop RLM as a systems house specialising in a mix of JORN and non-JORN systems work. The investment was justified partly by the volume of defence system work then in prospect and partly by Defence’s earlier policy of growing an Australian prime able to give substance to defence self-reliance by maintaining an in-country systems capability. In April 2003 RLM successfully completed JORN’s development and in May 2003 JORN achieved final acceptance by Defence. This marked the end of the completion phase of the JORN procurement process, some 17 years after the original announcement the Minister for Defence and at least 50% over the original budget agreed by the DAO and Telstra in 1991. The time taken and cost incurred in executing demand for JORN to this stage are summarised in Table 8.7.
332 Ibid., p. S137, para 36. 333 JCPAA Report No 357: op cit page 47 para 4.22. 334 Johnson, Interview 25 July 2005, op cit. 180
Table 8.7 Executing demand for JORN
Dates Activity Time elapsed Cost incurred
(years) ($Am)
Oct 1986 – Dec Preliminary 4 3 1990 design studies
(AWA, BHP, Telstra)
Jun 1991 – Feb JORN 6 741.1 (ceiling 1997 procurement price) (initial phase - Telstra)
Feb 1997 – Nov JORN 2 1999 procurement (transition phase – Telstra/RLM)
Nov 1999 – May JORN 4 945 (Sept 1997 2003 procurement prices/exchange (completion rates) phase – RLM)
8.6 Post-acceptance development of JORN This thesis uses post-delivery development of a military innovation as an indication of its value. The two specific indicators of such value used in the thesis are the willingness of the original customer to invest in ongoing development of the innovation and the willingness of new customers to purchase the innovation, adapted as necessary to their needs, thereby encouraging its diffusion. This section considers the post acceptance development of JORN from this perspective and concludes with a summary of the time taken and cost incurred in such development.
The announcement by the then Minister for Defence in October 1986 that the government intended to procure JORN initiated two separate streams of JORN-related development activity. The first stream entailed investment in ongoing development, primarily by DSTO, of 181
OTHR at the JFAS at an estimated cost of $57.5 million (1986 prices). Contracts for this first stream were awarded to AWA and CSA (with the former being purchased by British Aerospace in April 1996). The second stream entailed the design, development and production, initially by Telstra and later by RLM, of two radars that complied with JORN Operational Performance Directive. One of those radars was located near Longreach in central western Queensland and the other near Laverton, in the goldfields region of Western Australia. The rationale for this arrangement is explained in the following paragraphs.
The JORN Operational Performance Directive that drove the design, development and production of JORN was based on 1980s thinking and technology. Hence both the Defence customer and RLM recognised that the Longreach and Laverton radars that were the focus of the JORN project would need to be upgraded to evolving JFAS standards. The defence customer also recognised that certain features of JFAS would need upgrading before its integration into JORN as the final element of the three-radar system. JFAS upgrade included path-dependent improvement of technology. It also entailed improving the system’s operational utility by giving ADF personnel operational experience in OTHR management pending the full system’s coming on line. To this end, and in addition to the $57.5 million initially budgeted, the DAO awarded a series of supplementary contracts which, by 2007, had totalled $87.7 million. Such JFAS enhancements included software upgrades, introducing frequency agile beam scanning (primarily to reduce the risk of hostile jamming of the radar), upgrading of the original DSTO-designed ARO signal processor and of other elements of the frequency management system and finally the conversion of JFAS for remote operation from the JORN Coordination Centre, located at Edinburgh, near Adelaide.335
Included in the firm price of $945 million for the Longreach and Laverton radars was $145.5 million for initial maintenance and support by RLM for the period 2003-07. In addition, DAO awarded RLM separate contracts for JORN enhancements to bring the Longreach and Laverton radars up to JFAS standards. The initial enhancement contract of $65 million was awarded in 2004, followed by a further contract worth $18.45 million in 2006. In June 2007, the then Minister for Defence announced a further round of investment in JFAS and the two JORN radars to further integrate them into a single national OTHR capability, to improve the distribution of JORN-generated surveillance information to national agencies (notably Coastwatch) and to position the network for further development. This additional investment was shared by the two contractors, with RLM being awarded an additional contract for $262 million to upgrade the two JORN radars on this basis and British
335 Colegrove, Project Jindalee, pp. 829 182
Aerospace (who had purchased the AWA defence business in 1996) being awarded a contract for $131 million to upgrade JFAS on the same basis – see Table 8.8 below.336
The fusion of the JORN and JFAS radars into a single integrated system enabled that system itself to be integrated into the RAAF command and control system. This was achieved via the Vigilare command and control system which combines surveillance, airspace battle management and provision of a Recognised Air Picture (RAP) to higher Defence headquarters. The RAP combines data and information from over 250 sources in near real time. These sources include, in addition to JORN, the F/A-18 search and targeting radars and the Wedgetail AEW&C aircraft.
This post acceptance JORN development was scheduled to be completed in 2012, some nine years after JORN was accepted by the RAAF. Thereafter ongoing development of JORN for the Australian defence customer will depend on that customer’s perception of the capability value JORN contributes to a portfolio of air defence assets organised around a doctrine of multi-layered air defence of continental Australia and its northern maritime approaches. The first layer comprises the coarse coverage of Australia’s northern maritime approaches provided by JORN which provides a trip wire for the air defence system as a whole. The second layer comprises more localised coverage provided by AEW&C aircraft which, after being alerted by JORN, are deployed to generate the more precise target identification and location information needed to vector interceptors like the F/A-18 fighter aircraft. The third layer is provided by the search and targeting radars operated by the F/A-18 and, in due course, its designated successor, the F-35.
This doctrine of multi-layered air defence was not accepted by all actors in the customer element of the Australian defence competence bloc until the early 1990s. In 1997, at the same time RLM had succeeded in re-baselining the JORN project, the government authorised Defence to acquire four AEW&C aircraft (with options for an additional two aircraft) at an estimated project cost of $2.2 billion (December 1997 prices). In June 2004 the government approved the acquisition of an extra two aircraft for $A225.6 million. This increase in project scope combined with price and exchange rate variations to bring the estimated budget for the project to $A2.841 million (December 1997 prices)337 – approximately twice the cost of the JORN system. The US company Boeing supplied the aircraft, which includes an advanced Multi-role Electronically scanned Array (MESA) radar mounted on a dorsal fin on top of the fuselage. Boeing and its sub-contractors encountered technical difficulties that caused the project to run some five years late. Hence the full suite of AEW&C and OTHR-based surveillance capabilities was not achieved until 2013, some 27 years after the decision to acquire JORN in 1986.
336 Minister for Defence: Maintaining Australia’s Over-the-horizon radar capability, media release 064/2007 of 28 June 2007. 337 Australian National Audit Office (ANAO), ANAO Report No 20 2011-12, DMO Project Data Summary Sheets, pp. 201-203. 183
It is reasonable to conclude that the Australian defence customer will subordinate further investment In JORN to investment in development of the multi-layered air defence system as a whole. In these circumstances, ongoing development of JORN is likely to be path dependent and intermittent. Such development is unlikely to provide incentives for Lockheed Martin and BAE Systems to invest in future JORN-related capacity – a situation recently recognised by the Australian defence customer.338
The two JORN industrialists, Lockheed Martin and BAE Systems, also seem likely to have difficulty in finding non-Australian customers for JORN-related systems. These difficulties derive from, firstly the Australian government’s policy stance and from, secondly, the characteristics of OTHR itself. The Australian defence customer has indicated that, in order to protect the military advantage Australia derives from JORN-based surveillance of the nation’s northern maritime approaches, it does not envisage permitting the export of JORN in whole system format. This position has remained unchanged since at least 1987.
The impact of the Australian defence customer’s policy on JORN’s export prospects is reinforced by the physics of ionospheric refraction of HF radar signals which limit demand for OTHR-based solutions to requirements for broad area surveillance. OTHR is inherently vulnerable to the vagaries of the ionosphere which reduces its military utility, particular in polar orientations. As the location of the JORN radars suggests, ionospheric refraction of HF radar signals occurs over continental distances, which limits the number of countries able to exploit the phenomenon for surveillance purposes. While OTHR can locate a target to a ‘column’ of airspace with sides of the order of a few tens of kilometres, it is poorly suited to providing precise information about target height.339 Finally, at the frequencies used by OTHRs, targets like aircraft tend to be about the same size as a wavelength – typically about 20 m – meaning that such targets simply appear as a point on the radar display, preventing the operator from classifying or otherwise interpreting them. Therefore, demand by non- Australian customers for OTHR-based solutions to requirements for broad area surveillance is inherently limited.
To date, the post-acceptance trajectory of JORN development has been determined exclusively by the requirements of the Australian defence customer. It is reasonable to conclude that this is likely to remain the case. On this basis the development of JORN following its acceptance by the RAAF in 2003 has been confined to upgrade of the JFAS radar, upgrade the JORN radars at Longreach and Alice Springs, and integration of all three radars into a networked system. The cost incurred and time taken to complete these activities are summarised in Table 8.8.
338 See Defence Materiel Organisation, Priority Industry Capability Health Check 2013 – High Frequency and Phased Array Radars, available at www.defence.gov.au/dmo/id/pic/docs/PIC_FactSheet_HFPAR.pdf, accessed 14 June 2013. 339 Although use of multiple radars and locator beacons can improve this aspect of its performance – see J. Thomason, Development of Over-the-Horizon Radar in the United States, Proceedings of the International Radar Conference, Institute of Electrical and Electronic Engineers Conference Publications, 2003, p. 600. 184
Table 8.8 JORN post-acceptance integration and upgrade
Dates Activity Time Elapsed Cost incurred ($Am) (years)
Oct 1986 – Feb JFAS upgrades 21 57.5 + 87.7 = 145.2 2007 (AWA/BAE & CSA) 131 Jun 2007-12 JFAS integration 5 (AWA/BAE)
Feb 2004 – Feb JORN upgrades 3 65 + 18.45 = 83.45 2007 (RLM) 262 Jun 2007-2012 JORN integration (RLM) 5
Oct 1986-Jun Total JORN post- 21 145.2+131+83.45+262= 2007 acceptance 621.65 integration and upgrade (RLM, BAE)
8.7 Conclusion This chapter demonstrates that each element of the Australian defence sectoral innovation system influenced JORN innovation outcomes. Specifically, Australian defence institutions affected the search for and selection of the initial prime contractor, leading to an attenuated schedule. Fractured networks among the actors performing the customer function led to choices that increased costs and extended schedules. Deficient links among the domestic elements of the Australian defence competence bloc offset strong links with overseas innovators, leading to gaps in the domestic technology base and exacerbating the 185
adverse impact on JORN costs and schedule of other deficiencies in the Australian defence sectoral innovation system. The next step is to investigate more deeply how the various elements of that system actually affected the time taken to develop JORN, the cost incurred in doing so and the pattern of JORN development following its acceptance into Air Force service. This is the subject of Chapter 9.
186
CHAPTER 9: The Australian Innovation System and JORN Innovation Outcomes This chapter analyses how the distinctive features of the Australian defence sectoral innovation system influenced the performance of that system. The chapter does so by explaining how those features influenced JORN innovation outcomes in terms of the time taken to progress the JORN project, the cost incurred in doing so and the trajectory of JORN development after the RAAF had accepted it into service. Accordingly, the chapter begins with an analysis of how Australian norms affected JORN innovation outcomes. This is followed by a discussion of how such outcomes were affected by the distribution of OTHR- related competencies among Australian actors and by the effectiveness with which those actors were linked together in an OTHR-related competence bloc. After that, the chapter investigates how Australian military doctrine affected JORN choices. Then follows a discussion of the links between the Australian radar technology base and JORN development. The chapter concludes with an analysis of how the way the Australian customer executed the demand for a solution to its requirement for surveillance of the continent’s northern maritime approaches affected JORN innovation choices.
9.1: The influence of Australian institutions This section explores how JORN innovation outcomes were influenced by the norms of, firstly, collective security; secondly, defence self-reliance; and thirdly open and effective competition for defence business. Australia’s focus from the late 1940s to the late 1960s on forward defence as part of its participation in collective security arrangements reduced the incentive for Australian defence actors to develop a capability for broad area surveillance of the nation’s northern maritime approaches. This focus and the associated emphasis on interoperability with friends and allies undermined any incentive for Australian innovators and entrepreneurs to invest in surveillance-related innovation. It helps explain why, in the mid-1950s, Australian defence scientists abandoned their early OTHR work at about the same time that the US Naval Research Laboratory continued to investigate the feasibility of using OTHR to obtain early warning of Soviet airborne attack.
Even when OTHR began attracting local scientific interest, the Australian defence customer remained indifferent. This helps explain why, despite the enlightened support of Butement after 1959, Strath was forced to spend 10 years pursuing OTHR research on the margins of other defence scientific activity. This disinterest in the potential of OTHR persisted despite the Australian demand for more independent military capabilities that emerged towards the end of the 1960s. For example, Strath only managed to secure more Australian resources for OTHR research after 1968 because he could show that the US had not only succeeded in using ionospheric refraction of HF radar signals to track aircraft at long range but had also begun investing in operational systems.
187
While the ANZUS Treaty and institutions like the TTCP were undoubtedly necessary conditions for Australian access to highly classified US work on OTHR, they were not sufficient to provide Strath with automatic access to that research. A formal precondition for such access was DSTO’s status as a trusted recipient of US information, underpinned by a government-to-government security agreement. Less formally, Strath had to demonstrate to US decision-makers that Australia had accumulated sufficient OTHR knowledge to warrant US cooperation in further OTHR development on a bilateral basis. Despite the paucity of resources available to him, Strath succeeded in demonstrating that DSTO had accumulated sufficient OTHR knowledge to engage in a meaningful exchange with like- minded US scientists. Gaining access to US knowledge at this point helped set the trajectory of subsequent Australian OTHR development and prepared the way for the more substantive US input that affected the time and cost of Australian OTHR development.
The second Australian norm, that of defence self-reliance, was first publicly articulated in the Australian government’s 1976 Defence White Paper.340 The shift from ‘forward defence’ to ‘defence self-reliance’ resulted in more emphasis on capabilities for maritime surveillance, reconnaissance and offshore patrol, including affirmation of Australian sovereignty in Australian waters and its offshore maritime resources zone. This emphasis helped foster support by defence capability planners for niche development of OTHR in Australia and enabled Strath and his DSTO colleagues to sustain development of OTHR following the end of Project Geebung in 1972 through Jindalee Stage A and Jindalee Stage B and up to the decision to acquire JORN in 1986.341
To maintain support for niche development of OTHR, however, Strath and his colleagues had to demonstrate, quickly and cheaply, that OTHR had sufficient potential military utility to warrant key defence decision-makers funding OTHR and deferring funding for competing technologies like AEW&C aircraft. Overall, demonstrating the feasibility of using OTHR to monitor Australia’s northern maritime approaches incurred modest costs but did take considerable time. For example, despite US assistance, some 13 years elapsed from the first observation of ionospherically enhanced radar signals in 1959 to the end of Project Geebung in 1972. The Strath team’s ability to absorb that assistance efficiently and effectively depended on their extensive involvement in prior OTHR-related learning-by-doing. Overall this combination of US assistance and DSTO in-house expertise enabled Strath and his team to demonstrate the feasibility of OTHR within two years of the decision to proceed.
Australian defence scientists took a further 14 years and spent some $40.7 million in going through the three pre-procurement stages of the JORN program. Those three pre- procurement stages comprised, firstly, establishing the feasibility of using OTHR in Australia to track ships and aircraft (Jindalee Stage A, finishing in 1972); secondly, demonstrating the
340 Killen, Australian Defence, pp. 10-11. 341 ibid, p. 49, para 8. 188
military utility of OTHR (Jindalee Stage B, finishing in 1985); and thirdly, the decision to acquire the JORN (October1986). Importantly, development of the knowledge, individual skills and organisational competence required for this pre-procurement work was confined to DSTO and the commercially based innovators, entrepreneurs and industrialists who would be called upon to produce that network were, at best, peripherally involved.
In the first two pre-procurement stages, US assistance (provided on a government-to- government basis under ANZUS auspices) enabled Strath and his team to save time and money. The third stage involved the decision to abandon the minimalist option and to acquire a network of OTHR radars as announced by the Minister for Defence in October 1986. The third stage did not involve the US and was devoted to debate among the actors in the customer element of the Australian defence competence block about the relative merits of the OTHR and AEW&C surveillance paradigms. Resolving these differences probably added some two years to the JORN development and acquisition schedule. Of more longer term significance in terms of JORN innovation outcomes, however, was the Australian defence customer’s decision late in the third stage to use procurement of JORN to help establish in Australian industry the capacity to supply and support the materiel needed for the self-reliant defence of Australia. This decision bought into play the third norm, that of open and effective competition, in a way that tended to increase the time taken and cost incurred in designing, developing and producing JORN.
In Australian defence procurement, the principle of open and effective competition for Australian defence business was the foundation of institutional arrangements designed to improve the value for money obtained in the conduct of that business. In the case of JORN, there was never any question of optimising value for money via open competition for the business. The issue in JORN’s case was how to orchestrate effective competition for the business to achieve satisfactory value for money. With the benefit of hindsight, the institutional arrangements driving that process in the late 1980s seem to have been ill conceived and the actors involved ill informed. Ill-conceived institutional arrangements and ill-informed judgements by actors under the rubric of open and effective competition added to the time taken to design, develop and procure JORN. The provisions of the JORN contract, however, largely protected the defence customer from payment of the extra cost incurred. The DAO decision to select Telstra despite that company’s known deficiencies can be largely attributed to Telstra’s ill-judged willingness to accept virtually all cost and technical risk. The JORN Project Office’s efforts to protect the integrity of the JORN turnkey contract (which was negotiated within the framework of open and effective competition) by denying Telstra access to DSTO expertise seems equally ill judged. This is especially apparent when viewed in the light of RLM’s rapid turn-around of the JORN project after establishing integrated project arrangements that included DSTO. Such ineptitude on the part of DAO and Telstra added some three to four years to the JORN schedule. Again, however, the JORN target
189
price incentive contract capped the extra cost to the Commonwealth at the JORN ceiling price.
9.2: The influence of the Australian defence competence bloc This section analyses the efficiency and effectiveness with which Australian defence actors performed their competence bloc functions and how this influenced the time taken to generate JORN. This section focuses on how JORN outcomes were affected by the way Australian actors performed the competence bloc functions of informed customer, innovator, entrepreneur, venture capitalist and industrialist. Because the Australian defence sectoral innovation system was (and still is) demand driven, the Australian defence customer’s influence on JORN innovation outcomes was fundamental. However, the Australian defence customer’s influence on JORN innovation outcomes was affected by the interplay among the actors involved in the execution of the defence customer function. These included the Minister for Defence, capability planners (primarily members of the civilian force development and analysis organisation), enablers (primarily DSTO), procurers (primarily the DAO), and military users (primarily the RAAF). The following paragraphs analyse how, in performing the customer function in the Australian defence competence block, these actors influenced the time taken to develop JORN, the cost incurred in doing so, and the trajectory of JORN development after its acceptance into RAAF service.
In making and announcing the decision to acquire JORN in October 1986 the then Minister for Defence, the Hon Kim Beazley acted on the advice of defence planners. But as a reformer in a reformist government, Beazley was not a passive actor in the JORN decision- making process.342 He was a passionate advocate of defence self-reliance (which featured prominently in the platform of the Labor Party of which he was a member). He embraced JORN as one of several key investments in giving practical effect to defence self-reliance, ensured that his officials accorded the procurement of JORN commensurate priority and injected a sense of political urgency into that process.
Beazley also considered Australian industry involvement in the supply and support of Australian defence force materiel a key element of defence self-reliance. As a result, for example, he provided the political support needed to begin far-reaching reform and restructuring of the government factories and dockyards. He also endorsed the use of defence procurement of capital equipment to foster Australian industry capacity to supply and support that equipment. This extended to his endorsement of the JORN equipment acquisition strategy that underpinned the search for a JORN prime contractor initiated by Dr Malcolm McIntosh and pursued by McIntosh’s successors. In encouraging this strategy, however, Beazley took care not to influence the selection of a prime contractor for the JORN
342 see Andrews, The Department of Defence, p. 248. 190
project. In accordance with Australian governance, this function was performed by the head of DAO, exercising delegated ministerial authority.343
The next group of customer-related actors to influence JORN innovation outcomes were the capability planners. In the post Tange defence organisation, the capability planners were mostly civilian and primarily focused on managing overall force structure development. From this position they decisively influenced the development of OTHR up to the point at which the procurers began executing demand for JORN. The planners did so by protecting OTHR development during its vulnerable niche phase. They set the parameters for JORN procurement by determining the number, location and degree of integration of the radars. The planners also set the trajectory for JORN development by formulating (with input from DSTO and the RAAF) the demanding JORN operational performance directive. These parametric influences obviously affected the time taken to develop JORN and the cost incurred in doing so. More relevant for present purposes, however, is the way deficient links among planners, enablers, procurers and users exacerbated the impact of these technical challenges on JORN innovation outcomes.
Specifically, the priority accorded by capability planners to development and deployment of an OTHR-based capability for surveillance of Australia’s northern maritime approaches was not reconciled with the priority accorded by procurers to using OTHR-based innovation to grow an Australian-owned, Australian-controlled company able to support the self-reliant defence of Australia. Such policy misalignment helps explain the dismissal of Gilligan’s idea of Defence engaging GeneralElectric/TRW (who had built the USAF’s AN/FPS-118 Backscatter radar) to take the lead in the JORN project.344 This fundamental breakdown in the intra-customer network was only rectified when the DAO novated the JORN contract to a consortium led by Lockheed Martin in November 1999 – some eight years after the DAO and Telstra concluded the first JORN contract in 1991.
Within the customer element of the Australian defence competence bloc, DSTO performed the enabling function by providing OTHR-related scientific advice. DSTO was Australia’s sole repository of OTHR expertise before, and for several years after, the 1986 decision to acquire JORN. Pre-1986, DSTO reduced the time taken to develop JORN and the cost incurred in doing so by brokering access to US OTHR technology under TTCP and the ANZUS auspices. These formal structures legitimised the formation of dense interpersonal networks that enabled Strath and other DSTO actors to exchange OTHR-related information with their US interlocutors efficiently and effectively. Such exchanges reduced the time taken by, and the cost incurred by, capability planners in gauging the military utility of OTHR in the Australian context, particularly during the critical Jindalee Stages A and B.
343 The Hon Kim Beazley, interview, Canberra, 11 September 2007. 344 Testimony by M. Gilligan, p. 25. 191
After 1986, however, the positive impact on JORN costs and schedule of DSTO enabling was offset by the DAO’s interpretation of the risk-sharing arrangements embodied in the ‘turnkey’ contract concluded between DAO and Telstra. By precluding the establishment and operation of the requisite networks between DSTO and Telstra, the DAO helped cause delays to the JORN schedule that largely offset the reduced cost and compressed schedule achieved by DSTO through its OTHR-related collaboration with the US. The defence customer did not rectify this deficiency in the network between DSTO and DAO until 1997- 98 (some seven to eight years after the DAO and Telstra signed a target price incentive contract for JORN in June 1991). The requisite networks were only established after the institution of integrated project team arrangements involving DAO, DSTO and RLM (as the new the JORN industrialist). Deficient networks between DSTO, DAO and Telstra contributed substantially to the delays that characterised Telstra’s stewardship of the JORN contract.
As the procurement specialist in the customer element of the Australian defence competence bloc, the DAO affected JORN innovation outcomes by, firstly, the way it managed competition for the JORN contract and by, secondly, the way it applied the AOCI policy. The DAO’s insistence on an Australian prime contractor for JORN meant relegating the two most technically knowledgeable and managerially competent US companies (General Electric and Raytheon) to sub-contractor status. Effective competition for the JORN contract on this basis yielded a shortlist of three consortia (AWA/General Electric, BHP/Raytheon and Telstra/GEC-Marconi) of which two (BHP and Telstra) were headed by new entrants to Australian defence business and the third (AWA) had only limited exposure to development projects of JORN’s scale and complexity. These arrangements impeded access to the technical knowledge and managerial competencies required to undertake JORN and, in turn, contributed substantially to the cost and schedule overruns and performance deficiencies that characterised Telstra’s stewardship of the JORN project.
These difficulties were exacerbated by the DAO’s conflation of, on one hand, the objective of developing local defence industry capacity for defence self-reliance with, on the other hand, the objective of developing an OTHR-based broad area surveillance capability for defence self reliance. The conflation of these objectives combined with the DAO’s hands-off approach to commercial arrangements led to the selection of the Telstra/GEC-Marconi consortium as preferred prime contractor in December 1990 despite Telstra’s known lack of technical competence and GEC-Marconi’s paucity of OTHR-related expertise. By conflating industry policy and capability requirements and by adopting a laissez-faire approach to the organisation of JORN industrialist competencies, the DAO contributed substantially to the cost and schedule overruns and performance deficiencies that characterised Telstra’s stewardship of the JORN project.
The DAO also influenced JORN innovation outcomes by the quality of advice it provided to decision-makers regarding the capacity of competing tenderers to perform the industrialist
192
function of the competence bloc. DAO advice about the capacity of Telstra to supply a JORN that met requirements was coloured by Telstra’s ill-judged willingness to accept full cost and technical risk, irrespective of its capacity to perform the JORN task. Negotiating the turnkey contract with Telstra took six months (December 1990 to June 1991), during which the DAO seems to have spent less time and effort trying to devise procurement arrangements that maximised the likelihood of Telstra/GEC-Marconi delivering a compliant JORN and more time and effort trying to ensure that Telstra understood Defence’s requirements on a turnkey basis and to devise a contract that protected the Commonwealth’s financial interests should Telstra/GEC-Marconi fail to deliver a compliant JORN. Much of the schedule slippage and cost overruns that characterised Telstra’s stewardship of the JORN contract to the DAO’s poorly informed assessment of the willingness and ability of the Telstra/GEC- Marconi consortium to align their technical and managerial competence to the requirements of the JORN turnkey contract.
The DAO also influenced JORN innovation outcomes by the way it managed the turnkey contract with Telstra. Six years elapsed between the signature of the JORN contract (June 1991) and Telstra’s engagement of the Lockheed Martin/Tenix joint venture to manage the JORN contract (February 1997). During this period the DAO’s JORN Project Office interpreted the provisions of the turnkey contract so as to inhibit the exchange of OTHR- related information between DSTO and Telstra. By exacerbating Telstra’s already formidable difficulty in growing the competencies required to deliver a compliant JORN, the DAO contributed to the schedule delays, cost overruns and performance deficiencies that characterised Telstra’s stewardship of the JORN contract in this period.
Within the customer element of the Australian OTHR competence bloc, the military user role was performed by the RAAF. The RAAF supported higher priority for surveillance of Australian northern maritime approaches as defence self-reliance gained increased acceptance. But for much of the 12-year period between the establishment of Jindalee Stage A (April 1974) and the decision to acquire JORN (in October 1986), the RAAF’s senior leadership favoured an AEW&C solution to that requirement and was either agnostic about, or opposed to, the acquisition of an OTHR-based solution to Australia’s broad area surveillance requirement. RAAF ambivalence about OTHR offset the support by civilian capability planners and added at least a year to the schedule for OTHR development.
Throughout the 13-year period encompassing Jindalee Stage A, Jindalee Stage B and the JFAS test bed activity, the RAAF – along with the other services – was involved in debilitating intra-defence disputation. Such disputation within the customer element of the defence competence bloc reduced any incentive local companies might have had to invest in OTHR - related capacity in anticipation of defence exercising its demand for an OTHR-based solution. Hence, when Australian defence actors did turn to local companies for an OTHR- based solution to broad area surveillance after 1986, those companies took longer and
193
incurred higher costs in growing the requisite competencies than would otherwise have been the case.
In the seven-year period between the Minister’s decision to acquire JORN (October 1986) and the RAAF’s acceptance of JORN (May 2003), the RAAF was not directly involved in JORN development and production. During this period, however, the RAAF was heavily involved in the JFAS where, in addition to training future JORN operators, it participated in ongoing refinement of OTHR. Separately, it was involved in ongoing development of overall broad area surveillance capability by the acquisition of AEW&C aircraft. Both these activities influenced the future JORN development after its acceptance into RAAF service. After 1986, the RAAF used JFAS to begin the process of learning-by-using OTHR for surveillance of Australia’s northern maritime approaches. The results of this operational learning were fused with the results of continuing DSTO research and fed back into ongoing development of JFAS, sponsored by the RAAF. Such ongoing development of JFAS continued until the RAAF’s acceptance of JORN in 2003, after which emphasis shifted to, initially, bringing JORN as delivered up JFAS standard as developed and, subsequently, to integrating JFAS and JORN into a single system. This JORN upgrade and alignment was largely completed by 2007, four years after acceptance of JORN and at an extra cost of $83.45 million. The combination of RAAF learning-by-using, DSTO research and BAE Systems learning-by-doing helped define the post 2003 pattern of JORN development by defining the technological trajectory to be followed.
That trajectory was also influenced by the introduction of AEW&C aircraft into the RAAF’s portfolio of airspace surveillance and control assets. Procurement of AEW&C aircraft, long advocated by the RAAF, was foreshadowed by the Minister for Defence in 1986. The project to acquire AEW&C aircraft was approved by the government in 1997 (by which time the Lockheed Martin/Tenix joint venture had taken over management of the JORN project from Telstra). The RAAF accepted the six AEW&C aircraft into service in 2013, at an acquisition cost of nearly $4 billion and 27 years after the decision to acquire JORN in 1986. In doing so, however, the RAAF helped realise JORN’s full capability value by integrating it into a layered air defence capability comprising JORN for broad area surveillance, AEW&C aircraft for precise location and identification and F/A-18 fighters for intercept.
Overall, intra-customer network deficiencies related to both process and policy added between two and four years to the JORN development process. The Australian defence customer was shielded from the extra costs inherent in such delays, which were borne by suppliers as a cost of doing defence business. Such supplier-borne costs are likely to have been considerable: for example, by the time Telstra relinquished the JORN contract and quit defence business in 1999, it had incurred a $A606 million loss.
A distinctive feature of the Australian OTHR competence bloc is the way DSTO performed not only the enabler function within the customer element of that bloc, but also the 194
innovator function in that competence bloc as a whole. DSTO has performed the OTHR innovator function for well over 50 years. This role began with Strath’s opportunistic research (1958 to 1968). It continued with its investigation of ionospheric refraction of HF radio signals in the Australian environment (Project Geebung – 1970 to 1972). DSTO’s innovation then progressed from OTHR technology concept demonstrator in Jindalee Stage A (1974 to 1979) to prototype OTHR in Jindalee Stage B (1978 to 1987). Thereafter, DSTO’s innovation emphasised learning through the interaction of researcher, user and commercial supplier in the JFAS phase of OTHR development (1983 to at least 2012).
The longevity and continuity of DSTO involvement in OTHR-related research and development enabled DSTO to accumulate unparalleled knowledge of the characteristics of the Australian ionosphere, of the behaviour of HF radio signals in that ionosphere and of the management of three overlapping HF radar systems in broad area surveillance mode. DSTO’s ability to draw on this accumulated stock of knowledge reduced the cost, schedule and technical risk inherent in JFAS development and in the integration of the Longreach, Alice Springs and Laverton radars into a single integrated JORN system. This upside, however, was offset by other aspects of DSTO’s innovator role that caused the development of JORN to take longer and cost more than would otherwise have been the case. The competence with which the DSTO performed the OTHR innovator role reduced its incentive to engage the entrepreneurs and industrialists who would be asked to incorporate the OTHR technology into an artefact that complied with the JORN Operational Performance Directive. Hence those entrepreneurs and industrialists needed to acquire OTHR-related knowledge and to develop JORN-related competencies, causing JORN design, development and production to take longer and cost more than would otherwise have been the case. This downside of the DSTO role as innovator was exacerbated by its need to demonstrate sufficiently rapid progress in the niche development of OTHR to defer commitment to the rival AEW&C solution. This need for haste reduced the opportunity for DSTO to engage local entrepreneurs and industrialists in Jindalee Stage A and Stage B.
The importance of access to US OTHR research in enabling DSTO to make timely progress at reasonable cost has been noted already. Equally important to DSTO’s ability to perform the innovator function, however, was the plurality of ideas and diversity of experience afforded by the dense network linking DSTO scientists and their US counterparts. The two groups interacted intensively in the process of adapting US knowledge to Australian circumstances, including how to take advantage of Australia’s relatively benign ionospheric environment.345 The early establishment of this diversity of knowledge enabled DSTO to handle technological uncertainty as well as discontinuities like the abrupt winding down of the US programs more readily than would otherwise have been the case. DSTO’s intense networking with like-minded members of the defence science community at home and abroad contrasts sharply with its limited contact with entrepreneurs and industrialists who
345 J. Strath, interview, 21 November 2007. 195
would be required to exploit the scientists’ innovations by producing militarily competitive artefacts. Such insularity inhibited the development of the kind of industrialist competencies required to move from the Jindalee Stage B prototype to a system that complied with the JORN Operational Performance Directive.
DSTO’s ability to perform the innovator function was also vulnerable to dysfunctional institutional arrangements within the Australian defence sectoral innovation system. Hence, for example, the JORN Project Office’s interpretation of the risk allocation arrangements and associated liabilities between DAO and Telstra pre-empted the transfer of DSTO’s accumulated knowledge of OTHR to Telstra and helped lengthen the JORN schedule by several years. A concerted effort by RLM, new DAO executives and new JORN project managers was required to rectify these institutional deficiencies. RLM could not have achieved its subsequent and relatively rapid turnaround of the project without this initiative.
In the Australian OTHR context, the entrepreneurial function was performed by individuals whose knowledge enabled them to identify an opportunity, whose position in an organisation legitimised the action they took to pursue that opportunity and whose location in a network enabled them to marshall the resources required to realise that opportunity. JORN innovation outcomes were affected by the way in which, and the timing with which a number of individuals performed the entrepreneurial function. John Strath acted as a scientific entrepreneur in identifying the opportunity of using OTHR for broad area surveillance. Mike Gilligan acted as a capability entrepreneur in making the case for procurement of a network of OTHR, based on the pioneering work by John Strath. Malcolm McIntosh acted as a capability entrepreneur in taking advantage of the decision to procure a network of OTHR, as recommended by Gilligan, to foster an indigenous capacity in locally domiciled companies to support and upgrade JORN. Paul Johnson acted as a commercial entrepreneur in identifying the opportunity resulting from the technical, managerial and commercial difficulties encountered by Telstra and GEC-Marconi, and persuaded the principals of his company, Lockheed Martin, to take over and complete the JORN contract.
Strath recognised the potential utility of ionospheric refraction of HF radar because he combined long-established knowledge of the phenomenon, familiarity with more recent advances in HF radar signal transmission, reception and processing technologies and awareness of the Australian emergent demand for cost-effective surveillance of the nation’s northern maritime approaches. His position as a scientist in Australia’s fledgling defence research establishment helped legitimise his pursuit of the opportunity he identified. This aspect of Strath’s entrepreneurialism established the foundation for an OTHR-based solution to Australia’s broad area surveillance requirement.
DSTO’s membership of the TTCP gave Strath access to an international network of actors able and willing to contribute OTHR-related knowledge that complemented that 196
accumulated by Strath. This aspect of Strath’s entrepreneurialism enabled him to demonstrate OTHR’s military utility sufficiently quickly and cheaply to gain and retain support by capability planners for niche development of the technology. Between the start of Project Geebung (November 1970) and the announcement of the decision to acquire JORN 16 years later, Strath’s entrepreneurialism enabled DSTO to progress OTHR development sufficiently quickly and persuasively to demonstrate its capability value relative to AEW&C aircraft, the competing alternative advocated by the RAAF.
Gilligan, educated as a physicist, and initially employed as a defence operations research analyst, was able to recognise the JORN opportunity through a combination of knowledge of OTHR work by DSTO/Strath and awareness of the requirement for cost-effective surveillance of Australia’s northern maritime approaches resulting from the shift to defence self-reliance. He acted as a capability entrepreneur in formulating a strategic and operational case for development and acquisition of JORN that was sufficiently persuasive to displace the minimalist option then preferred. But Gilligan’s entrepreneurial influence within the customer element of the Australian defence competence bloc was constrained. It was sufficient to define the functional parameters within which demand for JORN would be exercised. But he was unable to determine legitimately how that demand would be met within the parameters he and other capability planners had set.
Defence had recruited McIntosh from the then Department of Industry, Technology and Commerce to formulate defence industry policy. McIntosh recognised the JORN opportunity through a combination of scientific training, knowledge of DSTO ‘s OTHR work and awareness of the demand for cost-effective surveillance of Australia’s northern maritime approaches. As head of the DAO during 1986-89, McIntosh acted as a capability entrepreneur in supporting efforts by the then Minister for Defence, Beazley, to foster Australian industry’s capacity to supply and support the materiel needed for the self-reliant defence of Australia, including broad area surveillance. In JORN’s case, McIntosh exercised delegated authority to influence which companies participated in the search for a JORN prime contractor and the terms and conditions on which they did so. McIntosh established the requirement for candidate JORN prime contractors to be Australian owned, Australian operated but teamed with overseas technology providers. While this established the basis for development of Australia’s indigenous capacity to supply and support JORN, it underestimated the time required to establish local competencies and, thereby, set the scene for subsequent delays and cost overruns. McIntosh left the DAO after shortlisting the AWA, BHP and Telstra bids in 1989. His successor presided over the selection of Telstra/GEC-Marconi in 1990 and the design of the turnkey contract in 1991.
Johnson led General Electric Aerospace marketing to the Australian Defence Organisation. His awareness of the JORN commercial opportunity dated from 1986-90, when he represented General Electric Aerospace in AWA’s unsuccessful bid for the JORN prime
197
contract. Telstra/GEC-Marconi’s mounting difficulties with the JORN contract (which became increasingly well known after 1994) gave him a chance to revisit this opportunity. Johnson began revisiting the JORN opportunity in 1996. In doing so he was able to argue that Defence needed to temper the importance it accorded to Australian control of JORN system design and software with recognition of the need for appropriate supplier competencies. On this basis he succeeded in brokering sufficient modification of the AOCI policy to allow Lockheed Martin to provide the technical and managerial leadership required to identify and eventually resolve the issues that caused the failure of Telstra/GEC- Marconi’s attempt to develop JORN. This due diligence process, which began in 1997 and ended with the novation by Defence of the JORN contract to RLM in 1999, set the parameters for successful completion of the project in 2003. A related aspect of Johnson’s entrepreneurialism was his ability to construct a commercial vehicle – RLM – that enabled him to deploy Lockheed Martin’s resources to retrieval of JORN while satisfying Defence insistence on Australian ownership and control of JORN technology. Johnson diagnosed, and devised a remedy for, the JORN project’s problems in two years (1997-99). He then led RLM in delivering a compliant JORN four years later (in 2003). Without Johnson’s entrepreneurialism, bringing JORN to a successful conclusion would have taken much longer.
The DAO performed venture capitalist functions in commissioning two years worth of risk reduction studies (October 1986 to December 1988) by selected Australian companies. Those companies undertook risk reduction studies related to the radar type, radar location, software development and project risk management. The studies were part of DAO’s management of effective competition. They helped inform DAO’s preliminary assessment of which combinations of company and technology warranted further consideration as potential JORN prime contractors and justified shortlisting three companies (AWA, BHP and Telstra) for the next round of DAO search for a JORN supplier. The DAO also performed venture capitalist functions in awarding contracts to the three shortlisted companies to undertake project definition study (PDS) contracts, to be completed between September 1988 and March 1989 at a total cost of $3 million. On the basis of the PDS, the DAO eliminated BHP and then invited AWA and Telstra to submit tenders in May 1990. On the basis of the tenders, the DAO selected Telstra as the preferred tenderer in December 1990.
The combination of risk reduction studies, project definition studies and tenders took a total of four years (1986-90) and probably cost Defence the equivalent of some 2% of the target price for JORN eventually negotiated with Telstra. However, the appeal to DAO officials of Telstra/GEC-Marconi’s misplaced willingness to accept virtually all JORN project risk completely overrode whatever information about the relative competencies of AWA/General Electric and Telstra/GEC-Marconi the DAO gleaned from the studies.
198
The industrialist task was begun by Telstra/GEC-Marconi in 1991, passed to the Lockheed Martin/Tenix joint venture in 1997 and finally completed by RLM in 2003, 12 years after the DAO and Telstra signed the original contract. The Lockheed Martin/Tenix joint venture was established as a vehicle for conducting the due diligence in preparation for a proposal to take over management of the JORN contract from Telstra. RLM reconfigured the JORN project and carried it through to a successful conclusion. Hence it was action taken by Telstra and RLM that affected the time taken to design, develop and produce JORN, the cost incurred in doing so and the pattern of JORN development after its acceptance into RAAF service.
Telstra’s ability to perform the JORN industrialist function was limited on both technological and managerial grounds. As the corporatised descendant of a government-owned, government-operated national telecommunications monopoly, Telstra’s technological competencies were concentrated in telecommunications. By 1990, when it was announced as the preferred tenderer for the JORN contract, however, Telstra’s competencies in telecommunications had begun to lag developments driven by microelectronics and computing. This lag was due to management decisions to run down in-house technological competencies, government policies hostile to expanding public ownership of the telecommunications sector and intense competition from commercial suppliers.346 As a result, Telstra’s technology base needed major enhancement and redirection to adapt to OTHR, let alone to meet the requirement specified in the JORN operational performance directive. Whatever ability Telstra had to adjust its technology base depended on the resourcefulness of its Research Laboratories. The latter dated from 1923 and had played a vital role during World War Two in developing, among other things, Australia’s radar technology base. In the post war era, however, Telstra and its predecessors looked to the Laboratories for overseas technology assessment and procurement advice. The Laboratories’ contribution to Australian telecommunications innovation became essentially derivative.347 Hence Telstra’s participation in the four years of risk reduction studies, project definition studies and tenders run by the DAO was totally reliant on its sub-contractor GEC- Marconi. This dependence did not change during Telstra’s stewardship of the JORN contract. Telstra also lacked the requisite managerial skills. This became apparent in Telstra’s dealings with its customer (the DAO) and with its main sub-contractor, GEC- Marconi. It was also manifest in its lack of basic project management competencies.
Prior to 1990, Telstra’s defence business had been confined to the provision of telecommunications for Defence exercises. As a government-owned monopoly, Telstra had neither opportunity nor incentive to develop the commercial skills required to negotiate effectively with the DAO. This lack of defence business acumen was manifest in its uncritical
346 For a compelling description of this policy contest see A. Moyal, Clear Across Australia: A History of Telecommunications, Nelson, Melbourne, 1984, pp. 358-363. 347 ibid, pp. 365-366. 199
acceptance of JORN project risk inherent in the ceiling price incentive contract for JORN and in its inability to access DSTO expertise. Telstra’s lack of defence business acumen contributed materially to the cost and schedule overruns that characterised its stewardship of the JORN contract.
Prior to 1990, Telstra had extensive experience in negotiating contracts involving the supply of advanced telecommunications technology with such overseas majors as Plessey and Ericsson. Because such negotiations involved path-dependent extensions of telecommunications technology with which the Laboratories were familiar, Telecom and its predecessors were able to negotiate as deeply informed customers. This was not the case with GEC-Marconi. Telstra’s inability to construct a commercially satisfactory relationship with GEC-Marconi was manifest in, for example, the latter’s tendency to produce software and hardware as specified, regardless of its fitness for purpose.348 Telstra’s lack of commercial skills contributed materially to the cost and schedule overruns that characterised its stewardship of the JORN contract.
The DAO awarded Telstra the prime contract in June 1991. Five years later (and with only a year to go before delivery was due), Telstra had made only limited progress in growing the competencies required to manage the prime contract. When Lockheed Martin/Tenix took over management of the JORN contract in February 1997, they found that, for example, Telstra had not agreed system specifications with Defence; that only 350,000 of 1.2 million lines of software had been written; and that Telstra had made no plan for system integration. Telstra’s lack of project management skills contributed materially to the cost and schedule overruns that characterised its stewardship of the JORN contract.
These difficulties were symptomatic of Telstra’s total inability to grow the competencies required to perform the JORN industrialist function in anything like the time required to achieve the revised JORN delivery schedule, let alone the original June 1997 delivery date. Telstra’s lack of JORN-related industrialist competencies probably added at least six years to the time taken to deliver JORN. Importantly, however, the terms of the target price incentive contract between the DAO and Telstra protected the latter from cost overruns beyond the target price. Under the terms of the contract, Telstra had to absorb such cost overruns, leading to its $606 million loss on the JORN project.
In 1997-98 RLM began reconfiguring the JORN project by, for example, deriving an agreed systems engineering requirements baseline for the system that would enable it to comply with the JORN Operational Performance Directive; establishing a plan for integrating and testing the system that was understood and agreed by the customer; establishing the Melbourne Integration Facility with the resources required to revise, develop and integrate software as necessary. RLM was no more able to grow these JORN industrialist
348 Joint Committee of Public Accounts and Audit, Report No 357, pp. 85-86. 200
competencies in the requisite timeframe than Telstra had been. Instead, RLM injected an experienced management team from Lockheed Martin and Tenix Defence Systems and introduced an Integrated Project Team organisation that pooled the knowledge of RLM, DSTO, the DAO and the sub-contractors. RLM also introduced project management arrangements more consistent with requirements of developmental projects of JORN’s complexity, including, for example, the co-location of development and integration teams, planned effort based on proven metrics, robust risk management processes and revised management reviews based on earned value management. In order to assemble a team with the appropriate technical resources, RLM took advantage of the wind down of US OTHR programs and seconded some 45 Lockheed Martin engineers to RLM for the JORN project. It took advantage of the exit from defence business of Telstra and GEC-Marconi and hand picked other engineers. Finally, RLM retained well-performing sub-contractors.
While RLM’s initiatives were not sufficient to reverse the schedule slippage that occurred during Telstra’s stewardship, they were sufficient to regain the customer’s confidence. By re-baselining the project, harvesting the gains that had been made by Telstra’s sub- contractors despite Telstra’s incompetence, and introducing rigorous project management procedures, RLM succeeded in delivering a JORN that complied with the JORN Operational Performance Directive in 2003, only six years after Telstra relinquished management of the JORN contract. In novating the JORN contract to RLM, Defence agreed to pay RLM an extra $136 million above the target price of $741.1 million (in April 1991 prices) agreed with Telstra. In return for this, RLM had to agree to turn JORN around on a fixed-price basis. It is a tribute to RLM’s industrialist competencies that it succeeded in producing a compliant JORN on a commercially viable basis within these stringent parameters.
9.3: The influence of Australian military doctrine Developing and maintaining an effective capability for broad area surveillance of Australia’s northern maritime approaches was a lower priority during the 1950s and 1960s when Australia emphasised forward defence. During 1958-68, therefore, Strath investigated OTHR on the margins of other activity, against the wishes of his immediate superiors and only because of the enlightened support of Buteman, the Chief Scientist. By the early 1970s, as Australian defence planners began to accord higher priority to surveillance of Australia’s maritime approaches, Strath’s work began attracting more interest. But so did competing proposals to acquire more capable long-range maritime patrol aircraft and the AEW&C aircraft being developed by the US. The resulting debate tended to be framed in terms of warning-time/lead-time doctrine. Proponents of defence self-reliance argued that Australia would have substantial warning of the emergence of the kind of major threat that warranted AEW&C capabilities. Early advocates of defence self-reliance saw the coarse coverage of Australia’s northern maritime approaches provided by OTHR as appropriate for the lower level threats considered credible in the shorter term and consistent with the warning that Australia was likely to receive for more demanding threats credible in the 201
longer term. Conversely, the RAAF was preoccupied with developing air defence capabilities more relevant to higher level contingencies. It argued that the lead times for development of AEW&C capabilities were longer than the proponents of OTHR thought. Accordingly the RAAF considered OTHR a dangerous distraction from the strategically more important task of acquiring an AEW&C capability.
The warning-time/lead-time doctrine tended to encourage the view that OTHR and AEW&C capabilities were mutually exclusive options for broad area surveillance of Australia’s maritime approaches. Proponents of OTHR were able to argue persuasively that investment in development of OTHR was more consistent with the prevailing warning-time/lead-time doctrine than acquisition of an AEW&C capability. This enabled them to protect the niche development of OTHR technology and to defer investment in AEW&C technology. This zero- sum approach to OTHR versus AEW&C options for broad area surveillance tended to be perpetuated as Australian doctrine began to place more emphasis on the geographic determinants of Australian security and as the notion of core force supplanted the previous notion of warning-time/lead-time balance in the late 1970s. Investment in OTHR was consistent with the kind of capabilities required to handle low-level contingencies that then commanded priority. Hence OTHR was a more appropriate candidate for inclusion in the core force than AEW&C aircraft which were primarily relevant to higher level contingencies. The latter, it was argued, could be acquired in a configuration and in numbers that were best determined in light of leading indicators that Australia expected to receive of the emergence of more substantial threat. The RAAF was unable to counter this argument at the time, in part because in the early 1980s it was preoccupied with absorbing the new F/A- 18 fighters and Orion P3 long-range maritime patrol aircraft. Core-force doctrine continued to favour investment in OTHR and discourage investment in AEW&C in anything but the longer term. As a result, the capability planners continued to support DSTO’s development of OTHR on a niche basis through Jindalee Stages A and B.
In his 1986 review of Australia’s defence capabilities, Paul Dibb articulated the connection between defence self-reliance and investment in Australian defence force structure in terms of layered defence capabilities. Dibb’s notion of a layered defence was subsequently endorsed, in slightly modified form, by the then Minister for Defence The Honourable Kim Beazley, in the concept of defence in depth, promulgated in the 1987 Defence White Paper. Central to both layered defence and defence in depth concepts was a capability for effective surveillance of the sea and air gap to Australia’s north. By the mid-1980s the JSET trials had provided a conclusive demonstration of the military utility of OTHR in the Australian environment. As a result, OTHR emerged as a key element in the new doctrine of defence in depth. As Beazley stated:
When the Government talks about a 10 year or a 15 year threat it is talking about an enemy capable of launching and sustaining an invasion aimed at seizing and holding a
202
very substantial portion of the country or overthrowing the Government – in short the sort of credible threat posed by Japan in 1942 … We can say that the sort of threat requiring total mobilisation of Australian resources and massive build up of the defence forces would take a regional power 10 to 15 years ... But what about a much lower level of threat which could deny us our ports through mining? Cutting our sea lanes, for example, interdicting our oil platforms or mounting harassment raids around our extended coastline? What sort of warning time do we have for these? … At the capability level we have a plethora of good intelligence on what our neighbours are buying, their training programs and so on. We can handle the 10 year stuff ... But at the low level threat level we are dependent upon political judgement and it may be wrong. We must know on a ‘time specific’ basis who and what is near and around our approaches. The only way to plug that hole is through a completely operational Jindalee ...349
The doctrine of defence in depth helped precipitate the decision to acquire a network of OTHR radars. Beazley’s announcement in 1986 initiated a process of JORN development and procurement that took 17 years to complete, in 2003, when JORN was finally accepted into RAAF service. However, the same doctrine also gave more weight to higher level contingencies than the previous doctrines structured around the concepts of warning time/lead time and core force. Acknowledgement of the relevance of higher level contingencies meant acknowledgement of the need to complement JORN’s comprehensive but coarse coverage of Australia’s maritime approaches with the more precise but inherently more circumscribed coverage provided by airborne microwave radars. Hence, in announcing the decision to acquire JORN, Beazley also indicated that:
the decision to proceed with OTHR development did not in any way pre-empt future decisions on airborne early warning and control (AEW&C) systems in the defence program. OTHR was needed irrespective of future decisions on AEW&C systems.350
The doctrine of defence in depth marked the end of the zero-sum perspective of the OTHR and AEW&C solutions and rendered AEW&C more competitive than hitherto. While this did not affect the trajectory of JORN development on the basis of the 1986 JORN Operational Performance Directive, it had major implications for the pattern of JORN development after its acceptance into RAAF service. From the mid 1980s through to the mid 1990s, Australia’s capability planners judged that Australia’s strategic outlook was sufficiently benign, and the prospect of higher level contingencies warranting AEW&C aircraft sufficiently remote, to allow a decision on AEW&C aircraft to be deferred. Priority was given to bedding down JORN and giving the RAAF time to acquire sufficient knowledge of the system to inform judgements about how many and what kind of AEW&C aircraft were required. By the late 1990s, however, the doctrine of defence in depth had been displaced by a doctrine of
349 Kim Beazley, Threats to the north – military and political (interview with Peter Hastings), Sydney Morning Herald, 5 March 1987, p. 13. 350 ibid., p. 3. 203
regional engagement. The latter doctrine interpreted defence self-reliance more expansively and placed more emphasis on not only defending Australia but also contributing the security of Australia’s immediate neighbourhood and supporting wider interests.351 This fostered debate about the number and characteristics of AEW&C aircraft needed to complement JORN. While investment in JORN continued, Australia also began laying the foundations for enhanced intercept capability. In 1997 the DAO sought proposals for AEW&C aircraft and in December 2000 it awarded the AEW&C contract to Boeing. By this stage Defence had novated the JORN contract to RLM (1999), the latter had brought the project under control and was on track to deliver a compliant JORN by 2003. A layered broad area surveillance capability comprising, among other assets, an upgraded and fully integrated JORN, six AEW&C aircraft and 70 F/A-18 aircraft all linked by the Vigilair command and control system was finally achieved in 2013. This was 27 years after the 1986 decision to procure JORN. Post 2013, the nature and scale of Australian defence investment in JORN development will be determined by the value it adds to this portfolio of surveillance assets.
9.4: The influence of the Australian technology base The protracted 13-year gestation of OTHR in the Australian context (1974-87) can be partly attributed to the perception by both capability planners and RAAF users of OTHR as an untried challenge to the established airborne microwave-based technological paradigm. In Levinthal’s model of technological speciation, the emergent technology (in this case OTHR) may eventually displace the initially predominant technology (in this case, AEW&C) if it offers sufficiently greater functionality to users able and willing to pay.352 During the 1970s, OTHR emerged as a potential challenge to the established airborne microwave-based technological paradigm espoused by the RAAF through the work by DSTO/Strath on the Jindalee Stage A technology concept demonstrator. Ten years later, during DSTO’s work on the Jindalee Stage B prototype OTHR-based surveillance radar (1978-87), OTHR was seen as a clear challenge to the RAAF’s airborne microwave-based technological paradigm In the Australian context, however, OTHR showed little sign of displacing AEW&C. By the early 2000s Australian defence capability planners and military users considered OTHR and AEW&C as mutually complementary technological paradigms and subsumed them both into an overarching doctrine of network-centric warfare.353
The technology base underpinning this composite broad area surveillance capability comprises both domestic and overseas (mainly US) elements. The domestic element of
351 J. Moore, Defence 2000: Our Future Defence Force, Commonwealth of Australia, Canberra, 2000, pp. 46- 52. 352 David Levinthal, p220. 353 See, for example, discussion of ‘Technology trends and the revolution in military affairs” in J. Moore, Defence 2000, pp. 107-110. See also T. McKenna, T. Moon, R. Davis and L. Warne, Science and technology for Australian network-centric warfare: function, form and fit, Australian Defence Force Journal, 170, March/April 2006, pp. 62-76. 204
Australia’s radar-related technology base is focused on supply and support of JORN. It was created as a deliberate act of policy by the customer element of the Australian defence competence bloc. The decision to do so was prompted in part by the failure of market mechanisms, mediated through the US defence export licence process, to meet Australia’s radar requirements. Supply and support of virtually all the other elements of the Australian broad area surveillance capability – including the microwave radars embedded in the AEW&C aircraft, in the F/A-18 aircraft and in the LRMP aircraft – is underpinned by the overseas element of the Australian radar technology base. As long as Australia continues to enjoy access to the US radar technology base on terms and conditions acceptable to the Australian defence customer, and for as long as the RAAF sees value in the comprehensive but relatively coarse coverage that OTHR provides, OTHR will remain a niche element of the Australian radar technology base.
The OTHR technological system comprises transmit and receive sub-systems. The time taken to develop JORN, the cost incurred in doing so and the pattern of post introduction development of JORN was largely determined by, firstly, the effort required to develop sub- systems and by, secondly, the effort required to integrate those sub-systems into a system- of-systems that complied with the JORN Operational Performance Directive. Sub-systems that lay on the critical path for JORN development included the HF signal generator able to produce signals of extremely high spectral purity, phased array technologies for forming and steering the OTHR beam and signal processing technology with extremely low internal electrical noise so as to detect the minute return signals reflected by the target and refracted back through the ionosphere. In 1986 the technological systems underpinning development of the above systems did not exist in Australia. The Australian defence customer’s decision to undertake the indigenous design, development and production of JORN in 1986 meant that innovators, entrepreneurs and industrialists had to establish these technological systems as well as the capacity to integrate them into an integrated system- of-systems. This took extra time and cost extra money.
GEC-Marconi was responsible for development of these sub-systems for JORN. The cost overruns, schedule delays and performance deficiencies that characterised Telstra’s stewardship of the JORN contract can be partly attributed to the difficulties GEC-Marconi encountered in the development of the HF radar transmit and receive sub-systems for JORN. As Britain’s main supplier of microwave radars for air, land and maritime operations, GEC-Marconi was to Britain what Ericsson was to Sweden. GEC-Marconi was generally competent in microwave radar technology. To build JORN HF radar transmit and receive sub-systems with the characteristics required to enable JORN to meet the JORN Operational Performance Directive, however, took the company beyond what it had previously attempted: Specifically, to achieve Operational Performance Directive-specified performance levels in detecting the weak signals reflected by targets, the internal electrical noise generated by the JORN radar receiver had to be reduced to unprecedented levels.
205
In addition, in order to maximise the freedom of choice in use of JORN operationally, the JORN Operational Performance Directive specified a degree of beam-steering agility that could only be met by an extremely flexible transmitter. To achieve this agility, the JORN transmitter had to be able to be set up for a new beam direction, a new frequency and a new waveform in microseconds. Achieving this degree of flexibility required GEC-Marconi to set aside its established expertise in analogue transmitter design and to adopt a digital design using software to achieve the requisite degree of transmitter agility.354 GEC-Marconi took some five years (1991-96) to develop these ancillary technologies. This was much longer than it had estimated would be required when Telstra and the DAO signed the JORN contract in 1991.
The adverse impact on the JORN schedule of GEC-Marconi’s technical difficulties was exacerbated by its managerial culture. Despite JORN being a developmental project, GEC- Marconi insisted on producing systems as specified by Telstra, not producing systems that worked. The impact on the JORN schedule of Telstra’s lack of technical and managerial competence was exacerbated by GEC-Marconi’s underestimation of the technical complexity inherent in developing critical enabling technologies and by GEC-Marconi’s managerial approach. Nevertheless, by February 1997, when the Lockheed Martin/Tenix joint venture took over management of the JORN contract from Telstra, GEC-Marconi had largely solved these technical difficulties. This meant RLM had a compelling incentive to acquire these ancillary technologies and to hire the personnel involved (particularly as they became available with the sale of GEC-Marconi to British Aerospace in January 1999). Timely access to these ancillary technological systems helped RLM complete delivery of JORN in 2003, only four years after Defence’s novation of the JORN contract to the company in November 1999.
9.5: The influence of Australian demand The DAO conducted its search for a JORN supplier in accordance with the Commonwealth procurement policies. Because of the importance and sensitivity of the OTHR technology, the search process was restricted to Australian nationals and to organisations complying with the AOCI provisions. Australian companies had neither opportunity nor incentive to grow OTHR-related knowledge and competencies beforehand. Hence application of the AOCI policy in the search for a JORN industrialist greatly increased the risk of cost and schedule overruns and performance deficiencies. The Australian defence customer accepted the risk in the interests of growing an Australian prime contractor with the technical knowledge, management skills and financial resources required to supply and support an artefact judged critical to the self-reliant defence of Australia. The DAO attempted to
354 Testimony by W. Bardo in Hansard record of hearings by the Joint Committee of Public Accounts (JCPA), review of the JORN Project, Canberra, 29 November 1996, pp. 4-5. 206
manage these risks by requiring the Australian primes to team with overseas companies with OTHR competencies. But the way the DAO orchestrated this teaming effectively excluded the two most knowledgeable companies from the process. Far from reducing the extra risk resulting from the AOCI policy, the search process in fact exacerbated the cost, schedule and technical risks involved.
The DAO structured its search for a JORN industrialist to maximise competition for the business within AOCI constraints. Partly for that reason, the search took well over a year. In May 1988 it invited companies to register interest and by September 1988 the DAO had contracted three shortlisted companies to undertake six-month Project Definition Studies, the output of which was incorporated in the request for tender. In May 1989 the DAO requested the three shortlisted companies to submit tenders for the project by August 1989, and in September 1989 the DAO convened the Defence Source Definition Committee to review the tenders. The companies shortlisted took another six months to complete the project definition studies (September 1988-March 1989) and then a further four months (May 1989-August 1989) to prepare their respective tenders. The process of running an effective competition for the JORN contract added some 18 months to the JORN schedule.
In preparing its tender, each tendering company was required to codify its bid in sufficient detail to permit the DAO to rank competing bids in terms of relative value for money. The process of selecting Telstra as preferred tenderer on this basis took over a year (September 1989-December 1990). In September 1989 the Defence Source Definition Committee began reviewing the tenders submitted by, respectively, AWA/General Electric, BHP/Raytheon and Telstra/GEC-Marconi. In early 1990 the Committee eliminated the BHP/Raytheon bid. By mid 1990 the DAO invited AWA/General Electric and Telstra/GEC-Marconi to submit revised tenders and in December 1990, on the basis of the DSDC recommendation, the Minister for Defence announced the selection of Telstra as the preferred tenderer, subject to the outcome of contract negotiations. The actual basis upon which the DSDC judged the relative value for money of the competing tenders is not on the public record. However, the broad judgements made by the Committee in choosing the Telstra/GEC-Marconi bid for the JORN contract despite that team’s known lack of OTHR knowledge relative to the AWA/General Electric team can be inferred from hearings conducted by the Joint Committee of Public Accounts and Audit from July 1996-March 1997. On this basis, judgements about the relative value for money of the bids by Telstra/GEC-Marconi and AWA/General Electric can be inferred. In the following paragraphs these bids are discussed in terms of their relative economy (that is, spending less on inputs), efficiency (that is, a measure of productivity or output relative to input), and effectiveness (that is, a measure of the impact achieved).355
According to the Joint Committee of Public Accounts and Audit, AWA submitted a bid that was some $52 million cheaper (equivalent to 8% of the ceiling price) than that submitted by
355 See http://www.idea.gov.uk/imp/core/page.do?pageId=1068398 accessed 14 November 2008. 207
Telstra. From the Defence customer’s perspective, however, the advantages of AWA’s more economical bid were offset by AWA’s greater aversion to cost risk: AWA capped its liability at $50 million above the ceiling price. In contrast, and from the Defence customer’s perspective, the disadvantages of Telstra’s less economical bid were offset by Telstra’s willingness to accept open-ended liability and to pay any cost overruns.356
In selecting Telstra, Defence accorded very high positive value to Telstra’s much greater willingness to accept JORN project risk and commensurately large negative value to AWA’s unwillingness to accept JORN project risk. Subsequent developments during Telstra’s stewardship of the JORN project suggest, however, that Telstra’s willingness to accept such risk, and the DAO’s confidence that the Defence customer’s interest had been protected were entirely misplaced. This misallocation of risk prepared the way for the cost overruns, schedule delays and performance deficiencies that were to characterise Telstra’s six-year stewardship of the JORN project (1991-97).
As already indicated, members of the Defence Source Definition Committee were well aware that, by virtue of its prior involvement in Jindalee Stage B, AWA knew more about OTHR technology than Telstra. Members of that Committee were also aware that AWA’s partner, General Electric, had accumulated deep knowledge of OTHR technology via its involvement in the USAF ‘s AN/FPS-118 program. Again, however, it would be simplistic to attribute the cost blowouts, schedule delays and performance deficiencies that characterised the first six years of the JORN project to the DAO’s failure to select the more technically proficient AWA/General Electric bid over the demonstrably less technically proficient Telstra/GEC-Marconi bid.
From the Defence customer’s perspective, the advantages of AWA/General Electric’s greater technical knowledge were partly offset by commercial difficulties then being experienced by AWA – see Chapter 8. In assessing the impact on JORN innovation outcomes of the decision to set aside the most technically proficient AWA/General Electric bid, it is necessary to recognise that Lockheed Martin subsequently bought the General Electric radar business, including its OTHR expertise. Lockheed Martin’s possession of this expertise was critical to the credibility of its review of Telstra’s management of the JORN project in 1995 and to its ability to take over JORN project management (in partnership with Tenix) in 1997. This level of expertise was also central to Defence’s willingness to give it (via RLM) full responsibility for the JORN contract in 1999 and to its eventual delivery of JORN in 2003. Against this background, the cost overruns, schedule delays and performance deficiencies that characterised at least the first six years of the JORN project (1991-97) were partly due to the arm’s length relationship between DAO and JORN industrialists. These arrangements resulted from the DAO’s interpretation of the norm of open and effective competition. Under the DAO’s interpretation of effective competition, it was appropriate to require
356 Joint Committee of Public Accounts and Audit, Report No 357, p. 36. 208
Australian companies to team with more knowledgeable overseas companies but inappropriate to intervene in the choice of partners and teaming arrangements. In hindsight, the less contrived arrangements proposed by Gilligan and actually executed by Johnson seem both more prudent and more efficient than the approach adopted by DAO. The Gilligan/Johnson model entailed engaging a knowledgeable foreign company as prime contractor, requiring that company to partner with a suitable local company, and requiring the partnership to work closely with DSTO. Under Johnson’s guidance, this model also achieved the aim of retaining JORN intellectual property in Australia and establishing the indigenous capacity to repair, maintain and adapt the JORN system. The way the DAO orchestrated the partnerships between local and overseas companies effectively denied the JORN project access to the most technically proficient industrialist for six years. It took action by a commercial entrepreneur to remedy the situation.
It would be hard to overstate the value accorded by Australian defence decision-makers in the early 1990s to exploitation by Australian-owned companies of OTHR technology for Australian defence purposes. This was reinforced by concern that, despite extremely generous support by the US Department of Defence for Australian OTHR research, the US government might not allow General Electric to transfer the technology that AWA required to produce an operational JORN system that met JORN Operational Performance Directive requirements.357 Australian concern over the release of US technology may have reduced the merits of the AWA/General Electric bid in the eyes of the Defence Source Definition Committee. The concerns would have become much more specific and compelling had AWA/General Electric been selected as preferred contractor and had their bid proceeded to actual contract negotiations.
JORN project schedule slippage and cost overruns can be attributed to the choice of Telstra/GEC-Marconi. At issue is the extent to which the above concerns about release of US technology caused the DSDC to rank Telstra/GEC-Marconi over AWA/General Electric in the final source selection. The available material suggests that, in fact, such concerns carried little weight, that the Defence Source Definition Committee gave more weight to concerns about AWA’s commercial acumen and that the Committee gave far more weight to Telstra/GEC-Marconi’s plan to use relatively novel digital technology rather than the conventional analogue technology in JORN signal processing. In hindsight, it is clear that the DSDC underestimated the cost, schedule and technical risk inherent in this technological choice.
Brennan (then the JORN Project Manager and a member of the Defence Source Definition Committee) told the Joint Committee of Public Accounts and Audit that Defence considered digital technology a better basis for long-term development of JORN than the analogue
357 Testimony by Ayers, p. 83. 209
technology underpinning the AWA/General Electric bid.358 This point was corroborated by Bardo, who was the responsible GEC-Marconi engineer, in his testimony to the same Committee.359
These views about the weight given to the effectiveness of GEC-Marconi’s digital technology were corroborated by Johnson. He observed that the AWA/General Electric bid may also have been unduly influenced by the approach that General Electric had used in the USAF AN/FPS-118 OTHR radar. According to Johnson, the AN/FPS-118 OTHR relied on massive signal power and commensurate computer power to process the return signals. This emphasis on raw power resulted partly from the need to accommodate the impact of the aurora borealis on the radar’s operation (a complication not shared by the equatorially oriented JORN). The General Electric approach was considered less elegant, more expensive and less consistent with the long-term growth envisaged for JORN than the DSTO solution.360 Members of the Defence Source Definition Committee accorded high value to the potential effectiveness of the Telstra/GEC-Marconi bid based on the presumption of Telstra access to DSTO OTHR expertise in the shorter term development of JORN, and perceived merits of GEC-Marconi’s digital technology in terms of JORN’s long-term development.
JORN innovation outcomes were influenced by the allocation of risk between prime contractor, customer and sub-contractors, the exchange of information between the prime contractor and the defence customer (including DSTO) and between the prime contractor and sub-contractors, and the creation and management by the Defence customer of RLM’s incentive to deliver JORN in accordance the revised contract. DAO procurement procedures encouraged the JORN Project Office to adopt an ‘arm’s length’ relationship with Telstra based on formal project reviews and periodic milestone payments. This arm’s length relationship was formalised in, and extended by, the turnkey contract based on price ceiling incentive payment arrangements concluded between the parties.
In retrospect, both the DAO and Telstra were naive in accepting the risk allocations inherent in a turnkey contract for a developmental project. In hindsight, the DAO assumption that Telstra’s open-ended acceptance of JORN project risk would protect the defence interest seems particularly ill considered. In effect, in return for what became a fixed-price contract, the DAO traded off JORN’s schedule for development and production of a broad area surveillance capability as specified in the JORN Operational Performance Directive.361 Telstra, for its part, was extraordinarily naive in pressing on with development of JORN despite its lack of OTHR competence and its inability to access DSTO expertise as a result of the JORN Project Office’s interpretation of the turnkey contract. As a consequence of the
358 Brennan: Testimony to JCPAA Hearings into JORN, 6 December 1996, op cit pp38-39. 359 Bardo, W.T.: Testimony to JCPAA Hearings into JORN, 29 November 1996, Hansard, page 3. 360 Johnson, P.: RLM involvement in supply and support of the JORN, interview of 25 July 2005. 361 Jones, G.: Testimony to Joint Committee of Public Accounts and Audit, Canberra, Tuesday 23 July 1996, Official Hansard Report, page 19. 210
above naivety on the part of both customer and supplier, Defence acquisition of a broad area surveillance capability slipped six years (from 1997 specified in the original JORN contract to 2003 when JORN was finally accepted into RAAF service), Defence abandoned its attempt to grow an indigenous prime contractor and Telstra exited the defence business, having incurred a loss of $A606 million.362
During Telstra’s stewardship of the JORN contract its primary means of communicating with the JORN Project Office was via monthly Cost Performance Reports and quarterly contract progress meetings. As the contract progressed and as difficulties mounted, Telstra’s Cost Performance Reports became progressively less clear in analysing the cause of variations and in explaining corrective action taken, and progressively less complete in their coverage of variations and in explaining the impact of each variation.363 Deficient Cost Performance Reports undermined communication between prime contractor and customer in quarterly progress meetings. Similarly, the arm’s length relationship between Telstra and JORN Project Office undermined the utility of technical reviews and resulted in Telstra progressing a non-compliant design for JORN.364 Communication between Telstra and the JORN Project Office slowed and eventually collapsed, resulting in much nugatory work by Telstra and its sub-contractors. For example, in its initial due-diligence review of the project prior to taking over the contract from Telstra, Lockheed Martin/Tenix found that only 50% of the software produced to that date was useable.365 Such nugatory work added directly to the slippage of the project and, indirectly, to the cost incurred by Defence when RLM took over the project.
Comparable communication difficulties developed between Telstra and GEC-Marconi. These difficulties derived in part from the complex distribution of responsibilities between Telstra and GEC-Marconi defined in the original sub-contract. Telstra was responsible for the overall design, GEC-Marconi was responsible for radar system design (including both software and hardware) while Telstra was responsible for installation and integration of those systems. The poor communications resulting from these cumbersome arrangements were exacerbated by Telstra’s poor systems engineering skills, resulting in poorly specified interfaces between systems and the lack of a coherent plan to manage and control integration. As a consequence, the commercial relationship between Telstra/GEC-Marconi eventually collapsed, causing further delays and higher costs to be borne by the contractors.366
RLM’s handling of information management demonstrates how the critical information- handling task might have been managed. It also demonstrates what impact better handling of information would have made in terms of project performance. On taking over
362 JCPAA Report No 357: op cit page 47, para 4.22. 363 McNally, Jindalee Operational Radar Network, pp. 16-17. 364 ibid., page 20. 365 Johnson, P. Interview of 25 July 2005. 366 JCPAA Report No 357, pp. 85-87. 211
management of the JORN project from Telstra in 1997, RLM accorded high priority to re- establishing communications with the Defence customer. As a first step, RLM introduced Integrated Project Team (IPT) arrangements with the express intention of improving communications among RLM, the Defence customer and the sub-contractors. RLM supported the IPT arrangements with other standard project management procedures that served to improve communications between the JORN customer and supplier. These included the introduction of a formal risk management process, the reinvigoration of program management reviews and the introduction of Earned Value Management disciplines.
RLM built its Melbourne Integration Facility for the express purpose of developing and integrating JORN software. This facility enabled the co-location of engineers responsible for JORN system development and system integration. Co-location of critical technical personnel and secure communication links to other stakeholders enabled RLM to avoid the difficulties created by the Telstra/GEC-Marconi arrangements. RLM’s communication- related initiatives were based on sound systems engineering practice. They were informed by the prior experience of some 45 engineers seconded from Lockheed Martin after the USAF wound down its OTHR program. They encompassed 150 engineers employed by the best software sub-contractor.367 RLM’s communication initiatives helped the company complete delivery of JORN by 2003, six years after taking over from Telstra (in 1997) and four years after it was novated the contract by Defence (in 1999).
The DAO’s attempt to use a price-ceiling incentive contract to influence Telstra’s behaviour failed in the face of Telstra’s technical and managerial incompetence and the ineptitude of the JORN Project Office’s handling of the JORN turnkey contract. In novating the JORN contract to RLM, the DAO reverted to the fixed-price contracting model with which it was familiar. As a consequence, the novated contract included a provision of $A136 million provision for risk – equivalent to some 18% of the JORN ceiling price ($741 million) originally agreed with Telstra.368 More generally, a fixed-price contract between the DAO and RLM was only feasible because of Defence’s willingness to abandon its previous arm’s length approach and to work with RLM to solve outstanding technical issues, because of the technical knowledge and managerial skill RLM was able to apply to the task and because of RLM’s detailed understanding of the task gained in the course of its technical review of JORN and of its two major due diligence studies. RLM also took advantage of GEC-Marconi’s hard-won success in designing and developing the digitised hardware that was critical to complying with the JORN Operational Performance Directive.
In retrospect, the DAO’s use of a turnkey contract to procure a developmental project from a contractor with known technical and managerial deficiencies was ill considered. With the
367 Johnson: Interview of 25 July 2005. 368 ibid. 212
benefit of hindsight, a more prudent and efficient approach would have been to divide the project into phases with separately contracted deliverables. This approximates what happened. For example, in re-baselining the project, RLM was able to take advantage of technical progress by GEC-Marconi and to build on the experience gained by Telstra sub- contractors. DAO’s ill-considered use of a turnkey contract model contributed to the cost overruns and schedule slippage that characterised Telstra’s stewardship of the JORN contract.
9.6: Conclusion The Australian customer did not value indigenous ‘make’ solutions to Australian defence capability requirements until the 1980s, relatively late in the Cold War period. This left Australian industrialists poorly equipped to take on a development project of JORN’s scale and complexity when called up to do so at relatively short notice. Accumulating the requisite knowledge and growing the requisite competencies took companies’ time and cost them money, which translated into JORN schedule slippage and cost overruns. The Australian defence customer’s belated attention to indigenous ‘make’ solutions as a result of the shift to defence self-reliance also left the DAO poorly equipped to handle procurement of complex developmental projects in general. But the DAO’s focus on procurement from off-shore companies and from local companies building offshore designs for most of the Cold War left it particularly poorly equipped to manage the untried local industrialists who responded to the DAO’s invitation to participate in the JORN project.
The DAO’s poorly developed procurement skills were manifest in its focus on protecting the Commonwealth’s financial interests at the expense of securing capability outcomes. Poorly developed procurement skills also help explain the DAO’s misplaced confidence that the target-price incentive contract provided sufficient incentives for Telstra/GEC-Marconi to grow the competencies required to deliver JORN in the time specified. Poorly developed procurement skills also help account for the DAO’s mishandling of the turnkey contract with Telstra/GEC-Marconi, including both ill-conceived allocation of risk between DAO and Telstra and denying Telstra/GEC-Marconi access to DSTO’s expertise.
The functioning of the Australian defence sectoral innovation system was also impeded by the DSTO’s role as both technology enabler and innovator in the Australian defence competence bloc. As the technology enabler within the customer element of the Australian defence competence bloc, DSTO was able to use its well-developed connection with the US defence science establishment and its formidable in-house capacity to expedite OTHR development in Australia. As OTHR technology innovator, DSTO was also able to pioneer OTHR development to the point at which key decision-makers accepted it as a viable element of Australia’s broad area surveillance capability.
In conflating the roles of enabler and innovator in the Australian defence competence bloc, however, DSTO denied Australian commercial entrepreneurs and industrialists both 213
opportunity and incentive to develop OTHR-related knowledge and competencies before DAO invited them to register interest in the JORN project. This exacerbated the general lack of knowledge and competence that characterised the innovator/entrepreneur/industrialist elements of the Australian defence competence bloc resulting from the customer’s lack of interest in indigenous ‘make’ solutions to requirements. As a result, innovations that depended on such knowledge and competence had to wait until commercial innovators, entrepreneurs and industrialists established them. The Australian defence competence bloc did foster both capability and commercial entrepreneurs. But the ability of those entrepreneurs to take full advantage of the opportunities they identified was inhibited by weak networks within and among all elements of the Australian defence competence bloc. Rectifying these network deficiencies entailed coordinating divergent interests and policies within the customer element, aligning customer and supplier incentives and developing complementary competencies. All this took more time and cost more money than would have been the case if the appropriate networks had already existed.
During the Cold War, the Australian defence innovation system was characterised by doctrinal instability and immaturity. The instability was most clearly manifest in the debilitating disputation within the customer element of the defence competence bloc over the relative merits of OTHR and AEW&C solutions to the requirement for a broad area surveillance capability. The immaturity was manifest in the time taken to see OTHR and AEW&C as two mutually complementary elements in an air defence system. Settling doctrinal instability (at least as it related to surveillance of Australia’s northern maritime approaches) and establishing a system perspective of broad area surveillance within the customer element of the competence bloc took time and, in turn, delayed investment in the industrialist competencies required to produce JORN.
The next step is to juxtapose the above analysis of what the development of JORN demonstrates about the performance of the Australian defence sectoral innovation system with the analysis undertaken in Chapter 6 of what the development of ERIEYE demonstrates about the performance of the Swedish innovation system. This is undertaken in Chapter 10.
214
Chapter 10: Explaining Divergent Innovation Outcomes – Swedish and Australian Case Studies Compared
The question addressed in this thesis was: Why do nations at comparable stages of economic development, with comparable political systems and with access to comparable technologies perform so differently in generating novel solutions to similar requirements for military capability? In addressing this question the thesis adopted a modified version of the sectoral innovation systems framework. This led to a series of subsidiary questions concerning the impact on innovation performance of specific features of defence sectoral innovation systems. These subsidiary questions were addressed in the Swedish case study (Chapters 4-6) and the Australian case study (Chapters 7-9). The case studies indicate that, during the Cold War period, radar-based innovation in Australia took much longer, cost much more, developed more narrowly and diffused much less than comparable innovation in Sweden – see Table 10.1.
TABLE 10.1 Comparing JORN and Erieye innovation outcomes
System (country) Time in Cost to point of Development Cost of development, to operation and diffusion development, operation diffusion AUD (Post acceptance) (Period, years) AUD
JORN (Australia) 1974-2003 = 29 1.24 billion Development for 0.621 billion369 domestic demand only (2003-2013)
Erieye (Sweden) 1985-1997 = 12 0.18 billion370 Development for 0.433 billion371 export only, diffusion to six countries (1998- 2013)
This chapter explains the divergent Swedish and Australian innovation performance manifest in Table 10.1 by synthesising the answers to the above subsidiary questions. This
369 See Chapter 8, Table 8.8. 370 Converted at 2012 exchange rate SEK1 = AUD0.1426. 371 See Chapter 5, Table 5.4 for estimate, converted at 2012 exchange rate SEK1 = AUD0.1426. 215
synthesis provides the basis for an integrated response to the question about differences in innovation performance posed at the outset of the thesis. The chapter is divided into four sections. The first section of the chapter compares the performance of the Swedish and Australian defence sectoral innovation systems as revealed in the case studies. The second section of the chapter draws attention to where this thesis has contributed to the literature on innovation in general and on military technological innovation in particular. The third section of the chapter analyses what the observed differences in innovation system performance suggest about the management of military technological innovation in small democracies. The fourth section of the chapter discusses what this analysis suggests about future directions of research and concludes the thesis.
Section 10.1: Comparing the distinctive features of the Swedish and Australian defence innovation systems In order to compare the performance of the Swedish and Australian defence sectoral innovation systems, this section is structured around the five defence sectoral innovation system building blocks described in Chapter 3. Hence the following discussion begins with a comparison of how Swedish and Australian institutions affected Swedish and Australian innovation performance. This is followed by a discussion of how the way Swedish and Australian actors performed the defence competence bloc functions affected innovation outcomes in each case. This leads to a discussion of the influence on Swedish and Australian innovation outcomes of their respective military doctrines. A discussion of the influence of the Swedish and Australian technology bases on each country’s innovation outcomes then follows. The section concludes with a discussion of how the way each country’s defence customer executed demand for a novel solution to their respective broad area surveillance requirements affected innovation outcomes.
10.1.1 Comparing Swedish and Australian institutions In this thesis the term institutions denotes those norms, rules or laws that regulate the relations and interactions between actors involved in the innovation process. The Swedish norms analysed in Chapter 4 comprised armed neutrality, corporatism and governance. The Australian norms analysed in Chapter 7 comprised collective security, defence self-reliance and open and effective competition for defence business. For comparative purposes, Swedish and Australian norms can be grouped under grand strategies, business paradigms and defence governance.
The norm of armed neutrality caused the Swedish defence customer to favour domestic ‘make’ solutions to evolving Swedish military requirements consistently over at least five decades. Consistent emphasis on ‘make’ solutions helped Swedish innovators, entrepreneurs and industrialists to accumulate and refresh technical knowledge and related competence in step with those evolving requirements. They could then draw on the stock of knowledge and competence they accumulated in one cycle of requirements to meet 216
demand for novel solutions in the next cycle of requirements. This helped Swedish innovators, entrepreneurs and industrialists to meet demand for novel solutions to the SwAF requirement for a rapid reaction surveillance capability in 12 years at a cost of $AU0.18 billion. This is less than half the 29 years and some 15% of the $AU1.24 billion it took their Australian counterparts to meet comparable Australian demand for a broad area maritime surveillance capability.
The norms of forward defence and defence self-reliance within the framework of the ANZUS alliance caused the Australian defence customer to favour ‘buy’ solutions for most of the Cold War period. This reduced the opportunity and incentive for Australian innovators, entrepreneurs and industrialists to develop the technical knowledge and business acumen required to meet the customer’s demand for novel solutions to Australian military requirements. When, during the decade of self-reliance, the Australian defence customer accorded higher value to domestic ‘make’ solutions to Australian military requirements for maritime surveillance, Australian innovators, entrepreneurs and industrialists had to grow or acquire the requisite knowledge and competence. In doing so they took longer and incurred higher costs than their Swedish counterparts in meeting comparable SwAF requirements.
Australian defence planners valued an OTHR-based solution to the requirement for northern maritime surveillance as pivotal to the credibly self-reliant defence of Australia. Australian defence procurers also envisaged using Australia’s OTHR-based IP to grow an Australian prime contractor able to contribute to the self-reliant defence of Australia. These perspectives of how best to exploit OTHR-based innovation confined the development of OTHR after JORN had been accepted into RAAF service to that required to meet RAAF requirements. Post acceptance development and integration of JORN was protracted and expensive: taken together, the upgrade and integration of the three JORN radars cost some $AU0.621 billion and was continuing in 2012, nine years after the original JORN was accepted into RAAF service in 2003.
Swedish defence planners valued a rapid reaction surveillance capability much more tactically. They viewed this capability as just one element of a densely integrated ground- based air defence system. That system comprised numerous assets with overlapping capabilities which provided considerable redundancy. Such redundancy meant that the security of the system was not entirely dependent on ERIEYE. The Swedish defence customer was therefore prepared to release ERIEYE for export. In doing so it benefited from non-Swedish customers prepared to fund the system’s ongoing development and, thereby, refresh Swedish knowledge and competencies. Swedish government support for ERIEYE export combined with Swedish Ericsson’s commercial incentives led to the international diffusion of ERIEYE after its acceptance into SwAF service. At about the same time Australia was upgrading and integrating the JORN system, Ericsson/SAAB might have spent the
217
equivalent of $AU0.433 billion (that is, 70% of the comparable Australian expenditure on JORN upgrade/integration) on adapting and producing 18 ERIEYE systems in varying configurations for six different countries. It is reasonable to conclude that, while Swedish grand strategy had little impact on post acceptance development/diffusion of ERIEYE, Australian grand strategy had a strongly restrictive impact on post acceptance development/diffusion of JORN.
This leads to discussion of the impact of distinctive Swedish and Australian defence business paradigms. During the four decades of the Cold War (1950-91), the seamless exchange of information between, on one hand, the Swedish defence customer and, on the other hand, Swedish commercially based innovators, entrepreneurs and industrialists meant that these commercial actors effectively shared in the execution of demand for Swedish military capability. Such Swedish corporatism meant that the Swedish customer channelled demand for novel solutions to Swedish military requirements to a stable cluster of Swedish innovators, entrepreneurs and industrialists, each of whom specialised in a particular area of technology. Swedish corporatism created a framework within which demand for novel ‘make’ solutions created a sustained incentive for specialist innovators, entrepreneurs and industrialists to refresh, update and align their knowledge and competence with that demand. Hence Swedish corporatism prompted a process of adaptive knowledge accumulation and competence development by Sweden’s specialised innovators, entrepreneurs and industrialists that helped them to meet evolving Swedish demand for novel solutions to Swedish military requirements with a speed and economy that their Australian counterparts were unable to match.
During the Cold War, the predominant Australian defence business paradigm was open and effective competition for defence business. The way the Australian defence customer applied this paradigm helped undermine incentives and reduce opportunities for Australian innovators, entrepreneurs and industrialists to accumulate the technical knowledge and develop the competencies required to meet demand for novel solutions to capability requirements. Hence, when the defence customer called on them to produce innovative solutions to the requirements for a broad area maritime surveillance capability, the stock of knowledge and competencies available to local suppliers ex ante was too shallow and patchy to produce those solutions ex post. Because local innovators, entrepreneurs and industrialists had to rectify this gap between available and required knowledge and competence, they took longer and incurred higher costs in producing a novel solution to Australia’s requirement for maritime surveillance than their Swedish counterparts took to meet the Swedish requirement.
By affecting both the opportunities and incentives for local suppliers to invest in the knowledge and competencies required to meet demand for novel solutions to military requirements, Swedish and Australian defence business paradigms helped cause the
218
performance of the Swedish and Australian defence innovation systems to diverge markedly in producing comparable innovations. Sweden completed ERIEYE development in less than half the time Australia took to develop JORN and spent the equivalent of 15% of the Australian outlay on JORN in doing so. That said, however, neither corporatism nor open and effective competition had much impact on, respectively, post acceptance development/diffusion of either ERIEYE or JORN.
The next institution to be discussed is defence governance. During the Cold War period, elected Swedish Ministers legitimised military capability requirements with only broad oversight by the Riksdag and with the aid of commissions. This helped to de-politicise Swedish defence innovation and to foster general acceptance of the need for the executive arm of Swedish government to address those requirements and of the parameters within which it would do so. During the Cold War, Swedish defence capability planners and defence procurers operated relatively autonomously within the broad framework of the elected government’s policy guidance and resource allocations. This loose executive accountability, Swedish corporatism and defence customer autonomy allowed all elements of the Swedish defence competence bloc to exchange innovation-related information efficiently and effectively.
By contrast, the Australian Parliament held the Australian Minister for Defence and other elements of the Australian defence customer group closely accountable for their innovation choices. Such tight accountability combined with the prevailing emphasis on open and effective competition to inhibit the exchange of innovation-related information between the customer and other elements of the Australian defence competence bloc. Information flows within the Australian defence competence bloc were constrained relative to those in the Swedish defence competence bloc. This helps account for the more protracted JORN development schedule (17 years from the inception of Project Geebung to the closure of Jindalee Stage B) relative to that of Erieye (15 years from the inception of the Erieye feasibility studies to the beginning of production for the SwAF).
10.1.2 Comparing Swedish and Australian competence blocs The following paragraphs analyse how Swedish and Australian innovation outcomes were influenced by the way actors performed the competence bloc functions of, respectively, the informed customer, the innovator, the entrepreneur, the venture capitalist and the industrialist. The case studies undertaken for this thesis suggest that the informed customer can influence the performance of defence sectoral innovation systems in several ways. These include how the customer promulgates requirements, the customer’s knowledge of the capacity of other elements of the defence competence bloc and the customer’s ability to make informed trade-offs in managing cost, schedule and technical risk.
219
The Swedish defence customer was informed at several levels, which combined to reduce the cost of Swedish innovation and to reduce the time taken to realise those innovations. The nature of Swedish planning enabled the Swedish customer to make, efficiently and effectively, an informed and internally consistent judgement ex ante about the value of a proposed innovation. In Swedish circumstances, such informed customer consensus helped foster a commercial environment that encouraged Swedish innovators, entrepreneurs and industrialists to invest in the knowledge and competencies required to generate novel solutions to requirements. This worked to reduce the time taken and the cost incurred by those Swedish innovators, entrepreneurs and industrialists in responding to the customer’s demand.
From the late 1970s on, the disputation within the customer element of the Australian defence competence bloc over, firstly general capability levels and over, secondly, the appropriate solution to broad area surveillance requirements prevented that customer making informed and internally consistent judgements about the value of OTHR. Such disputation inhibited investment by Australian innovators, entrepreneurs and industrialists in OTHR-related knowledge. As a result, they had to establish the knowledge and competence required to respond to demand for novel solutions to Australian requirements, thereby adding to the time taken and cost incurred in doing so.
In executing demand for novel solutions to Swedish military requirements, the Swedish customer was well informed about the knowledge and competencies of the relevant Swedish innovators, entrepreneurs and industrialists. This reduced the incidence of Swedish defence business mistakes and, to that extent, worked to reduce the time taken and the cost incurred by both customer and supplier in generating novel solutions to Swedish capability requirements. In Australia, the interpretation of the norm of open and effective competition that prevailed from the late 1970s to the early 1990s left the Australian defence customer less well informed about the capacity of candidate suppliers. This increased the incidence of business mistakes and worked to increase the cost incurred by, and the time taken by, both customer and supplier in meeting Australian demand for indigenous ‘make’ solutions in particular. In retrospect, for example, the selection of Telstra as the initial JORN prime contractor was an egregious business mistake that probably accounts for at least two and up to four of the 12 years it took to procure JORN (June 1991 to May 2003).
In Sweden, technically knowledgeable actors in the customer element of the Swedish defence competent bloc were densely networked with technically knowledgeable and managerially competent innovators, industrialists and entrepreneurs. Such dense networks enabled both parties to make informed trade-offs in managing the cost, schedule and technical risk inherent in devising novel solutions to Swedish requirements. Such informed trade-offs reduced the incidence of business mistakes and worked to reduce the time taken
220
and the cost incurred by both customer and supplier in generating novel solutions to Swedish capability requirements. In Australia, from the late 1970s until the early 1990s the defence customer had an ill-informed approach to risk sharing and failed to manage turnkey contracts to align producer incentives with customer requirements. The Australian customer’s deficient project management skills undoubtedly contributed to the protracted 12-year JORN procurement schedule (compared to the six years Ericsson Microwave Division took to deliver the six Erieye systems procured for the SwAF).
In Sweden, competent innovators existed in both the customer and industrialist elements of the Swedish competence bloc. But most Swedish innovators were located in the commercial firms that also husbanded the actors performing the entrepreneur and industrialist functions. Hence Swedish firms responding to SwAF/FMV demand for a rapid reaction response capability employed innovators who were relatively conversant with both old and new technologies relevant to the requirement. This helps explain the relatively compressed (12-year) schedule that characterises the Erieye innovation performance.
In the 1970s, the Australian innovators able to combine technologies to address new military capability requirements were concentrated in the government/customer element of the competence bloc. When Australian firms were asked to respond to the customer’s requirement for an OTHR broad area surveillance capability, in the 1980s, those firms had to recruit or grow actors with the relevant competencies. This took considerable time and money and helps account for the relatively protracted schedule (23 years) and higher cost ($A1.23 billion) that characterised the JORN innovation performance. During the Cold War, Swedish innovators maintained a stock of technological knowledge that was close to that required to devise novel solutions to Swedish requirements, thereby enabling them to adjust that stock of knowledge to meet new requirements more quickly and cheaply than their Australian counterparts. Conversely, Australian commercial innovators took longer and incurred higher costs when they were asked to devise comparable solutions because they had to grow the requisite knowledge.
Both Sweden and Australia fostered entrepreneurs with the competence required to judge, ex ante, what combination of technologies (and suppliers) was most likely to meet capability requirements. Each fostered capability entrepreneurs in the customer elements of their respective competence blocs and each fostered commercial entrepreneurs in the firms that designed, developed and produced solutions to capability requirements. But the intra- customer networks linking the Australian capability entrepreneurs were less effective than their Swedish counterparts in facilitating the exchange of information about requirements and possible solutions to those requirements. Less efficient and effective information exchange helps account for JORN’s relatively protracted schedule (23 years compared to 12 years) and higher cost ($A1.23 billion compared to $A0.18 billion) that characterised the JORN and Erieye innovation performances.
221
In addition, Australian commercial entrepreneurs were more institutionally constrained than their Swedish counterparts. In the latter case, both the SwAF and FMV provided Ericsson (and later SAAB) commercial entrepreneurs substantial support in marketing Erieye overseas. In contrast, efforts by Telstra and RLM to gain permission to market JORN overseas were denied by capability planners. This helps account for the divergent patterns of post acceptance development that characterised JORN (path-dependent development for Australian requirements only) and Erieye (varied development to meet the needs of diverse overseas customers) indicated.
The Swedish and Australian defence sectoral innovation systems differed markedly in the role played by venture capitalists. In Sweden’s case, during the Cold War, Wallenberg influenced Swedish defence innovation outcomes by his skilful selection of competent managers and securing their appointment to key positions in the Swedish companies that hosted Sweden’s innovators, entrepreneurs and industrialists. Because the Australian defence competence bloc lacked venture capitalists with anything like Wallenberg’s skill and influence, it is reasonable to conclude that Wallenberg helped, indirectly but nevertheless significantly, to account for Erieye’s truncated schedule, low cost and rapid diffusion relative to JORN.
The Swedish and Australian defence sectoral innovation systems differed markedly in the competence with which their respective actors performed the industrialist function. Prior to the development of Erieye, the Swedish firms involved had developed deep competence in the skills required to design, develop and produce novel solutions to the Swedish requirement for a rapid reaction surveillance capability. Development of deep competence was encouraged by Sweden’s long-standing, widely accepted and generally understood commitment to armed neutrality, the high value consistently accorded by an informed defence customer to indigenous ‘make’ solutions and an industrially competent venture capitalist. Such deep industrial competencies help account for the truncated schedule, modest cost and wide diffusion that characterised Erieye innovation performance.
An industrially competent venture capitalist, supportive government policy settings and a long history of competing successfully in overseas markets encouraged Swedish industrialists to look for opportunities beyond the Swedish defence market. Acceptance of Erieye by a Swedish customer known for its technical competence enhanced Erieye’s appeal to non-Swedish customers. These factors help account for the speed at which Erieye technology diffused and the diversity of customers buying it.
The environment for Australian defence industrialists was the converse: Australia’s commitment to ‘make’ solutions was contested, priority was accorded to industry capacity to repair, maintain and adapt military platforms and systems, and local firms had little incentive or opportunity to husband industrially competent actors. Hence local firms had to grow, almost from scratch, the competencies required to meet demand for the novel OTHR 222
solution to a broad area surveillance requirement. This helps account for the protracted schedule and relatively high cost that characterises the JORN innovation outcomes. In addition, Australian firms were prohibited from exporting whole JORN systems (for which demand was limited in any case). These factors help account for the lack of diffusion of JORN.
In summary, the way Swedish actors performed the various competence bloc functions, and the way they interacted in doing so, built, reinforced and perpetuated incentives and opportunities for them to innovate more quickly and cheaply and to invest more in post introduction development of novel artefacts than their Australian counterparts. Conversely, key Australian norms limited and constrained the development of incentives and opportunities for Australian defence actors to perform the various competence bloc functions and to interact in ways that reduced the time they took to innovate, that lowered the costs they incurred in doing so, and that prompted them to invest in post introduction development of novel artefacts. In summary, the actors and networks component of the respective Swedish and Australian defence innovation systems helped cause Australian defence technological innovations to take longer, cost more and develop/diffuse less than comparable Swedish innovations.
10.1.3: Comparing Swedish and Australian military doctrine During the Cold War, Sweden developed a dynamically stable military doctrine of ground- based air defence and an associated concept of operations centred on detection, identification and intercept of attacking Soviet aircraft and ships. The doctrine was functionally stable in the sense that, once it was established in 1950s, the fundamental tasks required to execute the Swedish model of ground-based air defence remained unchanged for the duration of the Cold War. This helped create a stable investment environment for Swedish innovators, entrepreneurs and industrialists. At the same time, however, Swedish doctrine was technologically dynamic. The portfolio of assets (including manned fighter aircraft, sensors and command and control systems) Sweden assembled to perform those tasks and execute that operational concept had to be developed and upgraded continually in order to keep that asset portfolio competitive with evolving Soviet capabilities. This resulted in a program of continuous investment that enabled Swedish innovators, entrepreneurs and industrialists to refresh and enhance their knowledge and competencies.
The SwAF’s requirement for a rapid reaction surveillance capability was the outcome of a stable doctrine of ground-based air defence and a consistently applied concept of operations. As a solution to this requirement, Erieye was an addition to an evolving portfolio of ground-based air defence assets that derived from the interaction of an informed Swedish defence customer with a cluster of specialised Swedish companies that husbanded knowledgeable and competent innovators, entrepreneurs and industrialists.
223
The stability of Sweden’s military doctrine and the consistency of the associated concept of operations encouraged Swedish innovators to continually refresh their technical knowledge related to the SwAF’s rapid reaction surveillance requirements, Swedish entrepreneurs to identify emergent opportunities to meet the SwAF’s requirement on a commercially viable basis and Swedish industrialists to invest in the capacity to design, develop and produce the requisite artefacts. The co-evolution of Swedish military knowledge and the Swedish defence competence bloc helps account for the relatively truncated schedule and modest cost that characterised the development, production and procurement of Erieye.
During the Cold War, Australian military doctrine was characterised by poorly focused concepts of operation and weak links between those concepts and capability development when compared to corresponding Swedish doctrine. Particularly in the mid 1980s, the concept of AEW&C-based broad area surveillance capability advocated by the RAAF was inconsistent with the emphasis placed by Australian defence planners on cost-effective surveillance of Australia’s northern maritime approaches. The concepts of core force and warning time proved impractical as tools for use by defence planners responsible for the development of Australian military capabilities in the post Tange defence organisation. Nor could those concepts inform judgements about the relative merits of ‘make’ versus ‘buy’ solutions to Australian requirements for novel military capabilities.
These characteristics of Australian military doctrine meant that, during the Cold War, Australian defence companies had little opportunity or incentive to foster innovators with knowledge about broad area surveillance related technologies, to develop entrepreneurs able to recognise and act upon broad area surveillance related opportunities and to prompt industrialists to invest in the capacity required to design, develop and produce broad area surveillance related artefacts. Hence it is reasonable to argue that, by adversely affecting investment behaviour by Australian companies, the characteristics of Australia’s military doctrine help account for the relatively attenuated schedule and high costs that characterised the JORN project.
10.1.4: Comparing Swedish and Australian technology The following paragraphs compare the Swedish technology base and the Australian technology base in terms of, firstly, their composition. The latter concept concerns the stock of technological systems populating the Swedish and Australian technology bases at a given time. The concept includes the depth of technologies available in the base, the plurality of techniques employed in the base and the diversity of applications for the technologies in the base. The second point of comparison relates to the alignment of the Swedish and Australian technology bases. This concept refers to the congruence or ‘fit’ between the extant technological base and that required to meet a defence customer’s demand for novel solutions to new requirements for military. The third point of comparison relates to the flexibility of the Swedish and Australian technology bases. This concept concerns the
224
amount of time required to, and the cost of, adapting the extant technology base to that required to meet a defence customer’s demand for novel solutions to new requirements.
The composition of the Swedish technology base was a product of the response by Swedish innovators, entrepreneurs, venture capitalists and industrialists to the Swedish defence customer’s requirement, sustained over some 50 years, for a militarily competitive ground- based air defence capability. Swedish demand for a neutral technology and preference for indigenous ‘make’ solutions encouraged Swedish innovators, entrepreneurs and industrialists to develop a deep technology base that encompassed research, development, design and production activities in each technological system. The composition of the Swedish technology base reflected the diversity of the portfolio of artefacts assembled by the Swedish customer to meet a ground-based air defence capability requirement that evolved over the 50 years of the Cold War. Hence the Swedish technology base encompassed the technological systems underpinning the design, development and production of, for example, airframes and aircraft propulsion systems, sensors, command control and communications systems and munitions.
By comparison, the composition of the Australian radar technology base in the late 1980s was thin and patchy. This was a consequence of the way the Australian customer formulated and satisfied, again over some 50 years, a requirement for a broad area surveillance capability in general, and a requirement for cost-effective surveillance of Australia’s northern maritime approaches in particular. From the 1950s until the early 1970s the requirement for northern maritime surveillance was met perfunctorily by long-range maritime patrol aircraft primarily intended for anti-submarine warfare. Australia imported from the US successive generations of these aircraft (and the systems embedded in them). They were repaired and maintained by the RAAF in-house. Hence supply and support of such assets had almost no impact on the local Australian technology base.
As the requirement for northern maritime surveillance became more compelling during the 1970s and into the 1980s, the customer initially framed the solution to the requirement in terms of the predominant military technological paradigm. This was microwave radars embarked on manned aircraft as manifest in AEW&C aircraft operated by the USN and the USAF. Although this paradigm became increasingly contested within the customer element as Strath and his colleagues succeeded in demonstrating OTHR’s potential military utility, neither the capability debate nor OTHR development work had significant effect on the composition of the technology base. As a result, when the Australian defence customer began searching for a local supplier of an OTHR-based solution to Australia’s broad area surveillance requirement, the requisite technological systems were virtually non-existent.
It is therefore reasonable to argue that depth and diversity of the Swedish technology base in being at the time Erieye was conceived contributed to the speed and economy with which that system was designed, developed and produced. It is also reasonable to conclude that 225
the shallowness and patchiness of the Australian radar technology base at the time the decision was made to procure JORN helps explain why designing, developing and producing JORN took much longer and cost much more than Erieye.
During the Cold War, the dynamic stability of Sweden’s ground-based air defence doctrine led the Swedish customer to demand novel, but path-dependent, solutions to evolving capability requirements. Such path-dependent demand combined with Swedish corporatism encouraged Swedish innovators, entrepreneurs and industrialists to invest in path- dependent development of the knowledge and competencies required to generate novel solutions to evolving requirements. For as long as Sweden’s ground-based air defence doctrine remained dynamically stable, such investments kept the composition of the Swedish technology base generally closely aligned with evolving requirements. While these conditions continued, Swedish innovators, entrepreneurs and industrialists could meet evolving requirements relatively cheaply and quickly.
Conversely, the Australian customer initiated the search for Australian innovators, entrepreneurs and industrialists to design, develop and produce an OTHR-based broad area surveillance system at relatively short notice. Hence the technology systems comprising the Australian technology base were poorly aligned to those required to absorb and adapt the DSTO-developed OTHR technology and to design, develop and build a system that complied with the JORN Operational Performance Directive. In these circumstances, and until those technological systems were established and aligned with JORN requirements, JORN production took longer and cost more than corresponding Erieye development.
A distinctive characteristic of the Swedish defence technology base, particularly towards the end of the Cold War, was the diversity and depth of the technological systems that populated it. The tendency of Swedish companies to specialise in particular technological systems meant that the diversity of technological systems was accompanied by commensurate diversity of corporate actors. Each of those corporate actors fostered deeply knowledgeable and competent innovators, entrepreneurs and industrialists. The upshot was a pluralist technology base able to respond efficiently and effectively to the customer’s demand for novel solutions to capability requirements. It is reasonable to argue that such pluralism combined with dense networks to enable the Swedish technology base to respond to Sweden’s demand for a rapid reaction surveillance capability more quickly and cheaply than would otherwise have been the case.
Conversely, during the design, development and production phase of JORN (beginning in the early 1990s and continuing until the completion of the JFAS integration in 2013), the indigenous element of the Australian radar technology base was focused on the relatively narrow cluster of OTHR-related technological systems. In 2003, when JORN was accepted into service, those technological systems were fostered by two companies, RLM and BAE Systems. As the JORN work wound down, both of these companies experienced difficulty 226
maintaining OTHR-related knowledge and competencies on a commercially viable basis prompting the Defence customer to intervene in attempt to preserve JORN-critical skills.372
In summary, the diversity of the technological systems comprising the Swedish ground- based air defence technology base, combined with the depth of the associated actors’ technological knowledge and competencies, enabled the Swedish defence technological base to generate novel solutions to both Swedish and non-Swedish requirements more quickly and cheaply than their Australian counterparts. Conversely, the shallowness and fragmented nature of the Australian broad area surveillance technology base and the narrowness of the technological knowledge and competence of the Australian actors caused the Australian defence technological system to take longer, incur higher costs and address a narrower range of development opportunities than corresponding Swedish programs. As a result, it is reasonable to conclude that the composition of the Swedish and Australian technology bases was a major cause of the divergence of the respective trajectories followed by Swedish and Australian defence technological innovation during the Cold War period.
10.1.5: Comparing Swedish and Australian demand The following paragraphs compare the influence on Swedish and Australian innovation outcomes of the processes by which they executed their respective demand for novel solutions to their respective requirements for a broad area surveillance capability. These processes entailed, firstly, searching for both a technical solution to their respective requirements for a broad area surveillance capability and for a supplier of that solution. Executing demand entailed, secondly, selecting both a technical solution and a supplier of that solution. Thirdly, executing demand entailed procuring the selected solution from the selected supplier.
The search by Swedish defence actors for a technical solution to the SwAF requirement for a rapid reaction surveillance system was path dependent in the sense that it entailed adapting the well-understood microwave pulse Doppler radar technology incorporated in the PS-46A radar. Swedish innovators combined this known technology with a relatively novel application of emergent electronically scanned array technology that they had been investigating for decades. Swedish corporatist norms and Ericsson Microwave Division’s proven competence in the design, development and production of radar meant that searching for a supplier was a trivial task for the Swedish customer, which did not affect search costs or schedule.
By contrast, the search by Australian defence actors for a technical solution to the requirement for a cost-effective capability for northern maritime surveillance entailed both
372 R. Wylie, Defence industry policy andiInnovation, Security Challenges Vol 9 (2), 2013, especially pp. 115- 116. 227
niche development of a relatively novel OTHR technology and displacement of the established airborne microwave radar technological paradigm. In addition, the search for a local industrialist able to design, develop and produce an OTHR-based system able to comply with the demanding JORN Operational Performance Directive was also disruptive in the sense that no local companies had the requisite knowledge and competencies, which they had to grow from scratch.
In summary, Swedish defence actors were able to conduct the search for both a technical solution to the SwAF requirement for a rapid reaction surveillance capability and a supplier of that solution more quickly and cheaply than their Australian counterparts were able to search for a technical solution to the requirement for northern maritime surveillance and for a supplier of that solution. This helps account for the difference between the Erieye and JORN cost and schedule outcomes indicated Table 10.1. On the other hand, the search process is unlikely to have affected the pattern of Erieye and JORN development following their respective acceptance into service.
The Erieye selection process involved a series of trade-offs. The management of these trade- offs was facilitated by the dense networks both within the customer element of the Swedish defence competence bloc and between that customer and the other elements of Sweden’s radar competence bloc located in Ericsson Microwave Division. The first trade off involved selection of the Erieye range, with the SwAF only prepared to pay for the minimum tactically effective range. The capability and commercial entrepreneurs succeeded in gaining acceptance of a much longer range. The second trade-off involved the selection of a platform on which to mount the Erieye transmitter and receiver systems. Swedish innovators and entrepreneurs readily accepted the SwAF’s arguments against mounting Erieye on a fighter and turned their attention to a commercial turbo-prop aircraft instead. The third trade-off involved the location of the radar systems operator. The SwAF eventually accepted the FMV capability entrepreneur’s arguments against locating the operator on board the aircraft. Overall, the various elements of the Swedish defence competence bloc managed the trade-offs involved in the Erieye selection process with a speed and economy that their Australian counterparts could not match.
Compared to the above Swedish selection process, the process by which Australian defence actors selected Telstra/GEC-Marconi as JORN suppliers was protracted and convoluted. The process was protracted by the DAO’s adherence to the prevailing norm of open and effective competition for a development contract. It was convoluted by the attempt to expedite a transfer of both technical knowledge and managerial competence to Australian companies with little or no prior experience with OTHR technology or broad area surveillance. It is therefore reasonable to conclude that Australia’s relatively protracted and convoluted selection processes made a significant contribution to the longer schedule for JORN development relative to that for Erieye.
228
The Erieye procurement process was a seamless extension of the Swedish search and selection processes described above. In particular, the detailed consultation that characterised the Erieye design process enabled customer and supplier to form a detailed, common understanding of the solution envisaged and the risks involved. This shared view enabled procurement to proceed on the basis of a straightforward fixed-price contract. When compared to the Erieye procurement process, the Australian procurement process for JORN was disjointed in the sense that the ineptitude that characterised the first phase of JORN procurement had to be rectified in a transition phase before the final (and ultimately successful) production phase could commence. Taken together, such disjointed procurement arrangements contributed strongly to the higher cost and more attenuated schedule that characterised the procurement of JORN relative to that of Erieye. That said, however, neither Swedish nor Australian procurement arrangements seemed to have affected the pattern of post acceptance development and/or procurement of either Erieye or JORN.
In summary, the processes by which Swedish defence actors executed demand for a solution to the SwAF requirement for a rapid reaction surveillance requirement combined with efficiency and effectiveness with which they exchanged procurement-related information enabled the Swedish defence technological system to generate novel solutions to Swedish requirements more quickly and cheaply than their Australian counterparts. Conversely, the time required to establish an OTHR technology paradigm and the restrictions and conditions imposed on the search for an Australian supplier caused the Australian search process to take longer and incur higher costs than its Swedish counterpart. This was exacerbated by the disputation associated with the selection of OTHR, by the flawed process used to select a supplier and by procurement-related incompetence on the part of both customer and supplier. As a result, it is reasonable to conclude that the processes by which Swedish and Australian customers exercised demand for a solution to their respective broad area surveillance requirements contributed significantly to the divergence of the trajectories followed by, respectively, Swedish and Australian defence technological innovation during the Cold War period.
Section 10.2: Contribution to military technological innovation literature The comparison of the structure, conduct and performance of the Swedish and Australian defence innovation systems undertaken in this thesis contributes to both the literature on military technological innovation and to the broader literature on innovation generally. It contributes to the literature on military technological innovation by demonstrating the analytical utility of the concept of sectoral innovation systems when applied, after modification, to the phenomenon of military technological innovation. Conversely, however, this thesis contributes to the mainstream innovation literature by demonstrating how generic concepts like sectoral systems of innovation can be refined and extended in the
229
process of applying them to analysis of the phenomenon of military technological innovation. These claims are amplified in the following paragraphs.
On one hand, the relatively large literatures dealing with military technological history and military capability development tend to be a-theoretical and to ignore the advances in broader innovation theory. There is also a relatively large literature on the organisational and institutional influences on military technological innovation which, however, tends to focus on such innovation in larger countries (notably the US) and which tends to ignore such innovation in smaller countries. The relatively modest literature on the economics of military expenditure tends to focus on either the macro-economic effects of such expenditure or on the economic and technological spill-overs from such expenditure. On the other hand, the mainstream innovation literature tends to undervalue military technological innovation as a source of empirical material relevant to broader research into innovation performance.
This thesis draws on the mainstream innovation literature to shed light on the phenomenon of military technological innovation. In particular, the thesis draws heavily on the systems thinking and actor/network logic that pervades much of the more recent literature that straddles the interface between evolutionary economics and innovation. The thesis contributes to both the mainstream innovation literature in general, and to the military technological innovation in particular, by applying concepts developed in the systems of innovation literature to military technological innovation. Specifically, the thesis demonstrates the analytical utility of Malerba’s notion of sectoral systems of innovation in framing the causal connections involved in military technological innovation in a way not found in the standard military innovation literature. But the thesis also contributes to the mainstream innovation literature by demonstrating how the generic concept of sectoral systems of innovation can be applied to military technological innovation to yield rich suggestions for further refinement and amplification of the concept.
This thesis used case studies to answer questions related to ‘why’ small nations pursue military technological innovation and ‘how’ they do so. In structuring the case studies, the thesis adopted Malerba’s building blocks (that is, knowledge, technology, demand, actors and networks and institutions) but adapted those building blocks to the requirements of defence sectoral innovation systems. In doing so the thesis broke new ground in formulating a methodology for undertaking case studies of military technological innovation. One key area of adaptation was the notion of ‘institutions’. Here the thesis modified Malerba’s conception of institutions as ‘rules of the game’ by extending it to include ‘norms’ which, in turn, included the ‘grand strategies’ by which nations seek to secure their security. The thesis also modified Malerba’s notion of ‘actors and networks’ by introducing the notion of a ‘defence competence bloc’ based on Eliasson’s more generic notion of a ‘competence bloc’. The third modification related to Malerba’s notion of ‘knowledge’ and entailed
230
focusing on that aspect of military knowledge relating to military doctrine, concepts of operations and force structure planning. Taken together, these modifications result in a methodology for case studies of military technological innovation based on the organising concept of defence sectoral innovation systems.
The divergent performance of the Swedish and Australian defence sectoral innovation systems was attributed to differences in the way the respective Swedish and Australian norms, actors and networks, knowledge, technology and demand influenced the time taken to develop and produce each artefact, the cost incurred in doing so and each artefact’s pattern of post acceptance development/diffusion. As Yin pointed out, identifying the patterns of causation observed in the case studies entailed making an inference – based on evidence collected during the case studies – that a particular event resulted from a prior event. Yin advocated testing the validity of any such inference by addressing rival explanations for the phenomena concerned. This testing is undertaken in the following paragraphs.
The starting point for testing the case study methodology used in this thesis is the observation that military technological innovation involves political decisions. Hence the differences in Swedish and Australian innovation performance could be attributed to patterns of behaviour explained in the social choice theory developed by Buchanan and Tullock.373 Social choice theory is based on the proposition that when people participate in political decisions they seek to maximise their respective individual utility and that such individual utility functions differ. In the context of military technological innovation, this proposition would suggest that, for example, Swedish corporatism was the product of Swedish firms’ ability to control the Swedish defence customer’s innovation choices to maximise the firms’ utility. Similarly, the social choice proposition would attribute, for example, the debilitating disputation within the customer element of the Australian defence competence bloc about how best to provide northern maritime surveillance to competing attempts by the RAAF advocates of an AEW&C solution and the central planning advocates of an OTHR-based solution to maximise their respective utility.
As the above examples suggest, social choice theory can help explain certain aspects of the conduct of defence sectoral innovation systems. But it is much weaker in explaining structural aspects of defence sectoral innovation systems like, for example, the origin of requirements for broad area surveillance capability. Social choice theory is also much weaker in explaining performance aspects of defence sectoral innovation systems like, for example, seeking access to US technology and absorbing/adapting that technology. The notion of defence sectoral innovation systems provides a more comprehensive, coherent
373 James M. Buchanan and Gordon Tulloch: The Calculus of Consent: logical foundations of democracy, University of Michigan Press, Ann Arbour, 1965. 231
and therefore persuasive explanation of divergent innovation outcomes than any of the other theories addressed in the literature review, including social choice theory.
Section 10.3: Managing military technological innovation This thesis has demonstrated the utility of a defence sectoral innovation system perspective to those practitioners seeking to improve innovation performance. The Australian case study illustrates the impact on innovation cost, schedule and diffusion of failure by the national defence customer to make choices informed by an adequate grasp of the interrelationships among institutions, actors performing defence competence bloc functions, military doctrine, the nature of the technology base and the execution of demand for novel solutions to requirements. Conversely, the Swedish case study illustrates how choices over time by the national customer can influence the performance of defence sectoral innovation systems by shaping institutions, influencing the performance of defence competence bloc functions, interpreting military doctrine, fostering the technology base and executing demand.
Within this overall framework, several themes related to the management of military technological innovation in small democracies emerged from the comparative case studies undertaken in this thesis. The first such theme concerns the way in which the customer formulates requirements for military capability and interacts with innovators, entrepreneurs and industrialists in executing demand for solutions to those requirements. The second theme concerns how actors performing the customer, innovator, entrepreneur and industrialist functions at the individual, group and organisation level learn, accumulate technical knowledge and husband innovation-related competencies. A third theme relates to the efficiency and effectiveness with which customers, innovators, entrepreneurs and industrialists exchange information about requirements, about demand for solutions to those requirements and about the trade-offs involved in meeting that demand for those solutions. A fourth theme concerns the allocation ex ante of innovation risk among customer, innovator, entrepreneur and industrialist and the allocation ex post of accountability for innovation outcomes among customer, innovator, entrepreneur and industrialist
10.3.1: Undertaking military technological innovation The case studies indicate that the structure, conduct and performance of defence sectoral innovation systems are fundamentally driven by, firstly, the requirements of a customer willing to pay and by, secondly, how that customer executes demand for a solution to that requirement. The Swedish case study suggests that the performance of local innovators, entrepreneurs and industrialists in providing novel solutions to a given requirement depends on, firstly, how closely related that requirement is to the stock of technical knowledge and innovation-related competence accumulated by those actors and on, secondly, the efficiency and effectiveness with which those actors can adjust their 232
knowledge and competence to meet the new requirement. The stock of technical knowledge and innovation-related competence maintained by innovators, entrepreneurs and industrialists populating a given defence sectoral innovation systems and the efficiency and effectiveness with which those actors can adjust that stock is largely determined by how the customer communicates requirements over time and how, over time, that customer executes demand for novel solutions to those requirements.
The Australian case study demonstrates that shifting grand strategies, inadequately articulated military doctrine and contested requirements reduce opportunities for innovators, entrepreneurs and industrialists to accumulate technical knowledge and innovation-related competencies and reduce the incentive for those actors to adjust the stock of technical knowledge and competence to align with evolving requirements. Overall, the thesis suggests that successful management of military technological innovation requires a consistent strategic rationale, a dynamically stable military doctrine and widely understood and accepted requirements for military capability.
10.3.2: Learning for military technological innovation The case studies indicate that the quality of learning at individual, group and organisational levels throughout the defence competence bloc influences the time taken to produce novel solutions to requirements and the cost incurred in doing so. The Swedish case study demonstrates how such multi-level learning influences the stock of technical knowledge and innovation-related competence available to the innovators, entrepreneurs and industrialists populating a defence sectoral innovation system, and thereby influences the efficiency and effectiveness with which they respond to the customer’s demand for novel solutions to a military requirement. The Swedish case study also shows how organisational learning influences the customer’s ability to judge, ex ante, which innovators are most likely to devise novel combinations of old and new technology that meet the requirement, which entrepreneurs are most likely to identify ways of exploiting innovations to meet requirements on a commercially viable basis and which industrialists can marshall the resource to exploit the opportunities identified by the industrialists. The Australian case study also highlights the vulnerability of learning at individual, group and organisational levels to norms and arrangements for executing demand that disrupt or inhibit such learning throughout the defence competence bloc. If, for example, the customer accords a low value to indigenous ‘make’ solutions to requirements then local innovators, entrepreneurs and industrialists will have little opportunity or incentive to invest in the organisational learning required to meet future requirements. Overall, the thesis suggests that the management of military technological innovation needs to give careful consideration to the opportunity and incentive for actors in all elements of the defence competence bloc to learn from business mistakes as well as business success.
233
10.3.3: Exchanging information for military technological innovation The case studies indicate that the conduct and performance of the defence sectoral innovation system is influenced by the efficiency and effectiveness with which actors performing the defence competence bloc functions exchange information about requirements, about executing demand for solutions to those requirements and about the trade-offs inherent in designing, developing and producing artefacts that meet requirements. The Swedish case study suggests that a defence sectoral innovation system can reduce the time taken to produce novel solutions to requirements and reduce the cost incurred in doing so by developing networks among the actors in the defence competence bloc that enable them to exchange uncodified information about requirements and solutions efficiently and effectively. Conversely, the Australian case study suggests that business norms and practices that inhibit and impede the exchange by members of the defence competence bloc of uncodified information about requirements and solutions will cause the development of novel solutions to requirements to take longer and cost more. Overall, the thesis suggests that managers responsible for military technological innovation who seek to minimise the time taken to produce novel solutions to requirements and the cost incurred in doing so need to ensure that the networks linking members of the competence bloc enable them to exchange uncodified information efficiently and effectively.
10.3.4: Allocating risk and accountability for military technological innovation The case studies indicate that the conduct and performance of a defence sectoral innovation system is influenced by the ex ante allocation of innovation risk among actors in the defence competence bloc and by the ex post allocation of accountability for outputs and outcomes among those actors. The Swedish case study suggests that customer willingness to accept upstream technology development risk and the customer’s ability to manage that risk through phased development in a technological niche establishes a common understanding of downstream artefact design, development and production risk among customer, innovator, entrepreneur and industrialist. That common understanding of downstream risk enables customer, entrepreneur and industrialist to make informed judgements about the appropriate allocation of artefact risk in negotiating procurement arrangements.
The Australian case study suggests that a tightly coupled chain of accountability for the provision of novel solutions to military requirements can create incentives for actors in the defence competence bloc to take actions that, in the first instance, inhibit innovation and that then cause the innovation that does occur to take longer and cost more than would otherwise have been the case. For example, tight accountability chains between procurers, Ministers and Parliament can inhibit innovation by reducing procurers’ willingness to fund
234
experiments, the failure of which can be construed as evidence of incompetence or improper use of public money. Conversely, the Swedish case study suggests that tight accountability between planners, procurers, Ministers and Parliament for capability outcomes but loose accountability for innovation outputs can encourage innovation in general and, by facilitating the interaction of all elements of the defence competence bloc, reduce the time taken to produce novel solutions and the cost incurred in doing so.
Section 10.4 Conclusion and directions for future research The above analysis highlights the pivotal importance of learning at the individual, group and organisational levels to the ability of defence sectoral innovation systems to generate novel solutions to military capability requirements in small democracies. Placing such multi-level learning at the centre of military technological innovation suggests several directions for future research in this area. If small democracies seek to use multi-level learning to help generate novel ‘make’ solutions to military capability requirements at acceptable cost, schedule and technical risk, such learning needs to be cumulative. The Swedish case study suggests that military technological innovation takes time – often decades – even in well- functioning defence sectoral innovation systems. Fostering cumulative multi-level learning in these circumstances requires sustained commitment to the innovation program and stable policy settings for management of the program. Such stability of requirements and of institutional arrangements can be difficult to achieve in small democracies. More research into the interrelationship between cumulative multi-level learning and dynamic stability of requirements and institutional arrangements seems warranted.
The above analysis also suggests that cumulative multi-level learning can help innovators, entrepreneurs and industrialists lower the cost schedule and technical risk inherent in designing, developing and producing novel solutions to military capability requirements. In capitalist democracies, however, many (if not most) innovators, entrepreneurs and industrialists are located in commercial organisations. To invest in cumulative multi-level learning, such commercial organisations need continuous flow of business generated by an informed customer. The Australian case study suggests that generating such steady business is at odds with the principles of open and effective competition. More research into the interrelationship between cumulative multi-level learning and competition policy seems warranted.
The above analysis suggests that, in order to generate insights that confer military advantage, cumulative multi-level learning must be focused. The Swedish case study suggests that, to achieve such focus, innovators, entrepreneurs and industrialists need to specialise in particular technological systems. But design, development and production of the artefacts used by small democracies competing for military advantage typically require inputs from numerous specialists. Hence more research into the interrelationship between, on one hand, cumulative learning by specialist individuals, groups and organisations and, on
235
the other hand, the networks through which such specialists exchange the fruits of cumulative learning seems warranted.
236
Bibliography Alexander, Arthur J.: The Linkage Between Technology, Doctrine, and Weapons Innovation: Experimentation for Use, RAND Corporation, Santa Monica, May 1981.
Alic, John A.: Trillions for Military Technology – How the Pentagon Innovates and Why It Costs So Much, Palgrave, New York, 2007.
Allison, David K.: New Eye for the Navy: The Origin of Radar at the Naval Research Laboratory (NRL Report 8466), Naval Research Laboratory, Washington D.C., 29 September 1981.
Andersson, B., J Nilsson and S-B Thelander: Radar PS-46A for Aircraft JA-37, Ericsson Review, Vol 60 (2), 1983 pp. 46-57.
Andrews, Eric: The Department of Defence (Volume V of the Australian Centenary History of Defence), Oxford University Press, 2001.
Australian National Audit Office: ANAO Report No 20 2011-12, DMO Project Data Summary Sheets, available at http://www.anao.gov.au/Publications/Audit-Reports/2012-2013/2011-12-Major- Projects-Report, accessed 20 August 2014.
Ball, Des: The strategic essence, Australian Journal of International Affairs, Volume 55, No 2, pp235- 248.
Bayerchen, A.: From radio to radar: interwar military adaptation to technological change in Germany, the United Kingdom, and the United States in Murray, W. And A. Millett (eds): Military Innovation in the Interwar Period, Cambridge University Press, Cambridge, 1996, pp 265-299.
Beazley, Kim: The Defence of Australia 1987, Australian Government Publishing Service, Canberra, 1987.
Beazley, Kim: Tenders Invited for Over-the-Horizon Radar Network, Ministerial Press Release no 113/89 of Thursday 1 June 1989.
Beazley, Kim: (Member for Brand): Hansard. House of Representatives, Canberra, Thursday 20 September 2007, pp. 37-44, available at http://www.aph.gov.au/Hansard/reps/daily/dr200907.pdf. Accessed 19 May 2009.
Bennett, P.H. and R.W. Cole: Study of Options for Jindalee/OTHR, Memorandum CDF 902/1985 - SEC685.1985 dated 20 October 1985 (Bennett) and 28 October 1985 (Cole), unpublished.
Bergek, A., S. Jacobsson, B. Carlsson, S. Lindmark and A. Rickne: Analyzing rhe functional dynamics of technological innovation systems: a scheme of analysis, Research Policy, No 37, 2008, pp. 407-429.
Bertola, P. and J. Teixeira: Design as a knowledge agent: how design as a knowledge process is embedded into organisations to foster innovation, Design Studies Vol 24 (2), March 2003, pp. 181- 194.
237
Bitzinger, Richard: Facing the future: the Swedish Air Force 1990-2005, RAND Corporation, Santa Monica, 1991.
Blondal, Jon: Budgeting in Sweden, OECD Working Papers, Vol VI, no 47, OECD, Paris 1998.
Brabin-Smith, R.: The Heartland of Australia’s Defence Policies, Working Paper No 396, Strategic and Defence Studies Centre, Canberra, 2005.
Brandstrom, Annika: Coping with a Credibility Crisis – The Stockholm JAS Fighter Crash of 1993, Swedish National Defence College, Stockholm, 2003.
Breschi, Stefan and Franco Malerba: Clusters, Networks and Innovation, Oxford University Press, Oxford, 2005.
Brown, Gary and Laura Rayner: Upside, Downside: ANZUS after Fifty Years, Parliament of Australia, Canberra, August 2001.
Brown, Louis: A Radar History of World War II, Institute of Physics Publishing, Bristol, 1999.
Buchanan James M. and Gordon Tulloch: The Calculus of Consent: logical foundations of democracy, University of Michigan Press, Ann Arbour, 1965.
Carlsson, Bo: Technological Systems and Industrial Dynamics, Kluwer Academic Publishers, London, 1997. Carlsson, Bo and R. Stankiewicz: On the nature, function and composition of technological systems, Journal of Evolutionary Economics, Vol 1, 1991, pp.93-118.
Coffey, Timothy, J. Dahlberg and Eli Zimet: The S&T Conundrum, August 2005, National Defense University, available at http://www.ndu.edu/CTNSP, accessed 14 June 2011.
Colegrove, S.: Project Jindalee: From Bare Bones to Operational OTHR, Proceedings of the Institute of Electrical and Electronic Engineers International Radar Conference, 2000 available at http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=851942&tag=1, accessed 20 August 2014.
Collis, B.: Fields of Discovery: Australia’s CSIRO, Allen & Unwin, Crows Nest, 2002.
Coulthard-Clark, C.: Breaking Free: Transforming Australia’s Defence Industry, Australian Scholarly Publishing, Melbourne, 1999.
Dahlsjo, O.: Antenna research and development at Ericsson, IEEE Antenna and Propagation Magazine, Volume 34 (2), April 1992, pp. 7-17.
David, Paul and Dominique Foray: Accessing and expanding the science and technology knowledge base in science, Technology and Industry Review, No 16, OECD, Paris, 1995.
Defence Materiel Organisation: Priority Industry Capability Health Check 2013 – High Frequency and Phased Array Radars, available at www.defence.gov.au/dmo/id/pic/docs/PIC_FactSheet_HFPAR.pdf, accessed 14 June 2013.
238
Defence Materiel Organisation: Defence Procurement Policy Manual: Mandatory Procurement Guidance for Defence and DMO Staff, Department of Defence, Canberra, 2013, available at http://www.defence.gov.au/dmo/gc/dppm.cfm, accessed November 2013.
Department of Defence: JORN Project Management and Acquisition Plan (Section 2), April 1993 in J.C. Gordon: JORN Papers for the JCPA, letter of 6 November 1996 to Secretary JCPA (REF DMD 96/31535 Pt 1) available in JCPA Submissions Exhibit No 9 (AR 28 OF 1995-96).
Department of Finance and Deregulation (Financial Management Group): Commonwealth Procurement Rules: Achieving Value for Money, Commonwealth of Australia, Canberra, 2012, p. 14, available at http://www.finance.gov.au/sites/default/files/cpr_commonwealth_procurement_rules_july_2012.p df, accessed November 2013.
Dibb, P.: Review of Australia’s Defence Capabilities, Australian Government Publishing Service, Canberra, 1986.
Dobinson, C.: Building Radar, Methuen, London, 2010.
Donovan, P.: Anticipating Tomorrow’s Defence Needs: A Century of Australian Defence Science, Commonwealth of Australia, Canberra, 2007.
Dorfer, I.: System 37 Viggen – Arms, Technology and the Domestication of Glory, Universitetsforlaget, Oslo, 1973.
Dosi, G.: Sources, Procedures, and microeconomic effects of innovation, Journal of Economic Literature, Vol 26, September 1988, pp.1120-1171.
Dosi, G.: The nature of the innovation process in G. Dosi, C. Freeman, R. Nelson, G. Silverberg and L. Soete (eds) Technical Change and Economic Theory, Pinter, London, 1988,.
Dosi, G and R. Nelson: An introduction to evolutionary theories in economics, Journal of Evolutionary Economics, Vol 4, 1994, pp. 153-172.
Edquist, Charles and Bjorn Johnson: Institutions and organisations in systems of innovation, in Edquist (ed.): Systems of Innovation: Technologies, Institutions, and Organisations, Pinter, London, 1997pp. 46-49.
Edquist, Charles: Systems of innovation: perspectives and challenges in J. Fagerberg, D. Mowery and R. Nelson (eds): The Oxford Handbook of Innovation, Oxford University Press, Oxford, 2006 pp. 181- 208.
Edquist, Charles (ed.): Systems of Innovation: Technologies, Institutions, and Organisations, Pinter, London, 1997.
Eisenhardt, Kathleen: Building theories from case study research, Academy of Management Review, Vol 14 (4), 1989, pp. 532-550.
239
Eisenhardt, Kathleen and M. Graebner: Theory building from cases: opportunities and challenges, Academy of Management Journal, Vol 50 (1), 2007, pp.25-32.
Ekeus, Rolf: Peace and Security: Swedish Security Policy 1969-1989, Abridged Version and Translation of SOU 2002:108, Statens Offentliga Utredningar, Stratsredsberedningen, Stockholm, 2004.
Eliason, Leslie C. and Emily Goldman: Theoretical and Comparative Perspectives on Innovation and Diffusion in Emily Goldman and Leslie Eliason (eds) The Diffusion of Military Technology and Ideas, Stanford University Press, Stanford California, 2003, pp. 1-30.
Eliasson, Gunnar: Competence Blocs in the Experimentally Organized Economy in Gunnar Eliasson (ed.): The Birth, the Life and the Death of Firms: The Role of Entrepreneurship, Creative Destruction and Conservative Institutions in a Growing and Experimentally Organized Economy, The Ratio Institute, Stockholm, 2005, pp. 27-125.
Eliasson, Gunnar: Advanced Public Procurement as Industrial Policy: The Aircraft Industry as Technical University, Springer, New York, 2010.
Ergas, Henry: The ‘New Face’ of Technological Change and Some of Its Consequences, unpublished memorandum, March 1994, cited by David and Foray in Accessing and Expanding the Science and Technology Knowledge Base in Science, Technology and Industry Review, No 16, OECD, Paris, 1995, page 43.
Evans, David: Air Power in the Defence of Australia: The Strategic Context, in Des Ball (ed.) Air Power: Global Developments and Australian Perspectives, Pergamon Press, Rushcutters Bay, 1988, pp.115- 131.
Fagerberg, J.: Innovation – a guide to the literature in J. Fagerberg, D. Mowery and R. Wilson (eds): The Oxford Handbook of Innovation, Oxford University Press, Oxford, 2005, pp. 4-5.
Finkel, M.: On Flexibility: Recovery from Technological and Doctrinal Surprise on the Battlefield, Stanford University Press, Stanford California, 2011.
Fleck, James: Configurations: crystallizing contingency, International Journal of Human Factors in Manufacturing, Vol 3 (1), 1993.
FOI Annual Report 2001, Stockholm, available at http://www.foi.se/upload/omfoi/infomaterial/foi- annual-report-2001.pdf, accessed 3 February 2010.
Foray, Dominique: The Economics of Knowledge, MIT Press, Cambridge Massachusetts, 2004.
Foss, N: The Emerging Competence Perspective in Foss, N and C. Knudsen (eds): Towards a Competence Theory of the Firm, London, Routledge, 1996.
Fredriksson, Urban: Airborne radars in the Swedish Air Force, available at http://www.x- plane.org/home/urf/aviation/text/airborne_radars.html, accessed 26 August 2011.
Freeman, C.: Technology Policy and Economic Performance – Lessons from Japan, Pinter, London 1987. 240
Freuling, Stefan: A History of Australian Strategic Policy since 1945, Defence Publishing Service, Commonwealth of Australia, Canberra, 2009.
Gilligan, M.F.: A Report to Secretary and CDF on Options for Over-the-horizon radar, Executive Summary, unpublished, Department of Defence, Canberra, May 1986.
Glete, Jan: Swedish Naval Administration 1521-1721: Resource Flows and Organisational Capabilities, Brill, Leiden, 2010.
Goebel, Greg: The SAAB JAS 39 Gripen, available at http://www.vectorsite.net/avgripen.html, p. 2, accessed 17 February 2010.
Goldman, Emily and L. Eliason (eds): The Diffusion of Military Technology and Ideas, Stanford University Press, Stanford, California, 2003.
Goldmann, Kjell: The Swedish model of security policy, West European Politics, Volume 14 no 3, 1991, pp. 122-143.
Granberg, Anders: Mapping the cognitive and institutional structures of an evolving advanced materials field: the case of powder technology, in Carlsson (ed) Technological Systems and Industrial Dynamics, Kluwer, London, 1997pp. 169-200
Gunnarsson, Gosta, Wilhelm Carlgren, Leif Leifland, Yngve Moller, Olof Ruin, Goran Rystad: Had There Been a War … .Preparations for the Reception of Military Assistance 1949-1969, Report of the Commission on Neutrality Policy (SOU 1994:11), Statens offentliga utredningar, Stratsredsberedningen, Stockholm 1994
Hacker, Barton C.: Military Institutions, Weapons, and Social Change: Towards a New History of Military Technology in Technology and Culture, Vol 35, no 4 (October 1994), pp768-834
Hall, Peter: Innovation, Economics and Evolution, Harvester Wheatsheaf, London, 1994
Hall, Peter and Robert Wylie: Isolation and technological innovation, Journal of Evolutionary Economics, Volume 24, 2014
Hall, Peter and Robert Wylie: Arms Export Controls and the Proliferation of Military Technology in Goldsmith, B. and J. Brauer (eds): Economics of War and Peace: Economic, Legal, and Political Aspects, Emerald Group Publishing, UK, 2010
Hammond, N.: Review of Auditor General’s Report 28, letter FASDM 456/96 of 27 November 1996, tabled as Exhibit No 17 in the JCPAA Review of the JORN Project
Hartley, Keith and Todd Sandler: Handbook of Defense Economics, Elsevier, 1995
Hastings, Peter: Threats to the North – military and political (Interview with The Kim C. Beazley), Sydney Morning Herald, 5 March 1987.
241
Headrick, J. and M. Skolnik: Over-the-horizon radar in the HF Band, Proceedings of the IEEE, Vol 62, No 6, June 1974.
Hitch, Charles J. and Roland N. McLean: The Economics of Defense in the Nuclear Age, Harvard University Press, Cambridge, 1967.
Holley, I.B. : Ideas and Weapons, Archon Books, Hamden, Connecticut, 1971.
Hurley, D.J.: The Foundations of Australian Military Doctrine, Australian Defence Publishing Service, July 2012, Canberra, page 3.2, available at http://www.defence.gov.au/adfwc/Documents/DoctrineLibrary/ADDP/ADDP-D- FoundationsofAustralianMilitaryDoctrine.pdf accessed 9 July 2014.
Industry Involvement and Contracting Division, Department of Defence: Review of the F/A-18 Industry Program, March 1994.
James, Andrew D.: Organisational change and innovation system dynamics: the reform of the UK government defence research establishments in Journal of Technology Transfer, Volume 34, 2009, pp505-523.
James, Andrew D.: Re-evaluating the role of military research in innovation systems in Journal of Technology Transfer, Volume 34, 2009, pp 449-454.
Jane’s Avionics online data base: Erieye AEW&C Airborne Early Warning & Control mission system radar, available at http://www4.janes.com/subscribe/jav/doc_view. accessed 29 August 2011.
Jane’s C4I Systems, 5 July 2011: TARAS Swedish Air Force communication system, available at http://jc4i.janes.com/docs/jc4i/search/results, accessed 11 August 2011.
Jane’s C4I systems, 28 July 2011: STRIL/STRIC air defence system (Sweden), available at http://jc4i.janes.com/docs/jc4i/search/results, accessed 11 August 2011.
Jane’s defence and security intelligence and analysis data base, available at https://janes.ihs.com/CustomPages/Janes/Home.aspx, accessed 21 August 2014.
Joint Committee of Public Accounts: Report No 243: Review of Defence Project Management, Volume 2 (Project Analyses), Australian Government Publishing Service, Canberra, 1986.
Joint Committee of Public Accounts and Audit: Report No 357 - The Jindalee Operational Radar Network, Commonwealth of Australia, Canberra, 1998.
Killen, D.: Australian Defence, Australian Government Publishing Service, Canberra, 1976.
Kinnaird, M., L. Early and B. Schofield: Defence Procurement Review 2003, Canberra, 2003, page 19, available at http://www.defence.gov.au/publications/dpr180903.pdf, accessed 18 March 2014.
242
Levinthal, David: The slow pace of rapid technological change: gradualism and punctuation in technological change, Industrial and Corporate Change, Vol 7 no 2 1998, pp 219-247.
Lundval, Bengt-Ake (ed.): National Systems of Innovation: Towards a Theory of Innovation and Interactive Learning, Pinter, London, 1992.
Lassman, Thomas C.: Sources of Weapon Systems Innovation in the Department of Defense: The Role of In-house Research and Development, 1945-2000, Centre of Military History, United States Army, Washington D.C., 2008.
Lyons, John, Richard Chait and Duncan Long: Critical Technology Events in the Development of Selected Army Weapons Systems: A Summary of Project Hindsight Revisited, Centre for Technology and National Security Policy, National; Defense University, Washington D.C., September 2006, available at http://www.ndu.edu/ctnsp/publications.html, accessed 3 June 2011.
McIntosh, M.: A New Environment for Industry, Defence Industry and Aerospace Report, Vol 9, no 11, June 1990.
McIntosh, M.: Letter, reference DEPSEC A&L 33/1990 of 8 February 1990, unpublished.
McKelvey, Maureen: Using Evolutionary Theory to Define Systems of Innovation, in Edquist (ed.) Systems of Innovation: Technologies, Institutions and Organisations, Pinter, London, pp 201-220
McNally, R: Jindalee Operational Radar Network (Australian National Audit Office Performance Audit No 28 1995-96), Australian Government Publishing Service, Canberra, 1996.
Marklund, Goran: STI OUTLOOK 2002 – Country Response to Policy Questionnaire (Sweden), page 9, available at http://www.oecd.org/science/inno/stioutlook2002nationalstpolicies.htm accessed 3 Feb 2010.
Malerba, Franco: Sectoral Systems of Innovation: Concepts, Issues and Analyses of Six Major Sectors in Europe; Cambridge University Press, Cambridge, 2004.
Markowski, Stefan, Peter Hall and Robert Wylie: Procurement and the Chain of Supply in Markowski, Stefan, Peter Hall and Robert Wylie (eds): Defence Procurement: A Small Country Perspective, Routledge, 2010.
McKenna, T.,T. Moon, R. Davis and L. Warne: Science and Technology for Australian Network-Centric Warfare: Function, Form and Fit in Australian Defence Force Journal, Issue 170, March/April 2006, pp. 62-76.
McNeill, William: The Pursuit of Power – Technology, Armed Force, and Society since AD 1000, The University of Chicago Press, 1982.
Mellor, D.P.: The Role of Science and Industry, Australian War Memorial, Canberra, 1958.
243
Metcalfe, J.Stanley.: Evolution and Economic Change in Ulrich Witt (ed.): Evolutionary Economics, Edward Elgar, Aldershot (UK), pp367-398.
Metcalfe, Stanley J.:, Evolution and Economic Change, in Witt, Ulrich (ed.), Evolutionary Economics , Edward Elgar, Aldershot, 1993.
Meurling John and Richard Jeans: The Ericsson Chronicle, Informationsforlaget, Stockholm, 2000.
Ministry of Defence: This is the Ministry of Defence, pp18-19, available at http://www.sweden.gov.se/content/1/c6/10/66/84/36f8afde.pdf, accessed 1 February 2010.
Mokyr J.: The Lever of Riches, Oxford University Press, New York, 1990.
Molas-Gallart, Jordi: Government defence research establishments: The uncertain outcome of institutional change in Defence and Peace Economics, Volume 12, no 5, 2001, pp417-437.
Molas-Gallart, J. and P. Tang: Ownership matters: Intellectual Property, privatisation and innovation in Research Policy, Volume 35, 2006, pp200-212.
Moore, J.: Defence 2000: Our Future Defence Force, Commonwealth of Australia, 2000.
Moores, Simon: ‘Neutral on our side’: US Policy towards Sweden during the Eisenhower Administration, Cold War History, Vol 2 no 3 (April 2002).
Morton, P.: Fire Across the Desert – Woomera and the Anglo-Australian Joint Project 1946-1980, Australian Government Publishing Service, Canberra, 1989.
Mowery, D. and N. Rosenberg: The Influence of Market Demand Upon Innovation; A critical review of some recent empirical studies, in N. Rosenberg (ed): Inside the Black Box: Technology and Economics, Cambridge University Press, Cambridge, 1982, pp193-241.
Moyal, A.: Clear Across Australia : A History of Telecommunications, Nelson, Melbourne, 1984.
Nelson, Brendan (Minister for Defence): Maintaining Australia’s Over-the-horizon radar capability, media release 064/2007 of 28 June 2007.
Nelson, R. (ed): National Systems of Innovation: A Comparative Study, Oxford University Press, Oxford, 1993.
Nelson, R and S. Winter: An Evolutionary Theory of Economic Change, Harvard University Press, Cambridge, 1982.
Nilsson, Mikael: Tools of Hegemony: Military Technology and Swedish-American relations 1945-1962, SANTERUS Academic Press, Sweden 2007.
Olsson, Ulf: At the Centre of Development: Skandinaviska Enskilda Banken and its predecessors 1856- 1996, Skandinaviska Enskilda Banken, Stockholm, 1997.
Olsson, Ulf: Furthering a Fortune Marcus Wallenberg Swedish Banker and Industrialist 1899-1982, Ekerlids Forlag, Stockholm 2001. 244
Olsson, Ulf: The Creation of a Modern Arms Industry: Sweden 1939-1974, Goteborg, 1977.
Pelz, Donald C., Fred C. Munson and Linda Jenstrom Dimensions of Innovation, Journal of Technology Transfer, vol 3 No 1, 1978 pp36-37.
Pierre, Jon: Legitimacy, Institutional Change and the Politics of Public Administration in Sweden, International Political Science Review, Vol 14, no 4, 1993.
Posen, B.R.: The Sources of Military Doctrine – France, Britain and Germany Between the World Wars, Cornell University Press, London, 1984.
Potts, Jason: The innovation deficit in public services: The curious problem of too much efficiency and not enough waste and failure, Innovation: Management, Policy and Practice, Vol 11, Issue 1, April 2009.
Potts, Jason: The New Evolutionary Microeconomics: Complexity, Competence and Adaptive Behaviour, Elgar, Cheltenham, UK 2000.
Pugh, Cedric: Sweden’s approach to improve effectiveness in Public Administration 1967-86, Financial Accountability and Management, Volume 4, N0 1, Spring 1988.
Ray, Robert: Government Approves Construction of Jindalee Over the Horizon Radar Network, Ministerial Press Release no 201/90 of 20 December 1990.
Richelson, J.: America’s Secret Eyes in Space: The U.S. Keyhole Spy Satellite Program, Harper Collins, 1990.
Rogers, Everett M.: Diffusion of Innovation (Fourth Edition), The Free Press, NY, 1995.
Rosenberg, Nathan: Inside the black box: Technology and Economics, Cambridge University Press, Cambridge, 1982.
SAAB Brochure: ARTHUR Weapon Locating System available at http://www.saabgroup.com/Global/Documents%20and%20Images/Land/ISTAR/ARTHUR/ARTHUR% 20ENG%20print.pdf, accessed 29 August 2011.
SAAB: PS-05/A Advanced Aviation Multi-mode Radar, available at http://www.saabgroup.com/Global/Documents%20and%20Images/Air/Sensor%20Systems/PS%200 5_A/PS05_100422.pdf, accessed, 28 August 2011.
Saxenian, A: Regional Advantage; Culture and Competition in Silicon Valley and Route 128, Harvard University press, Cambridge MA, 1994.
Schot, Johan and F. Geels: Niches in evolutionary theories of technical change: A critical survey of the literature, Journal of Evolutionary Economics, Volume 17, 2007.
Sherwin, Chalmers and Raymond Isenson: Project Hindsight: A Defense Department study of the utility of research, Science, Volume 156, June 1967, pp1571-1577, available at http://www.sciencemag.org/content, accessed 3 May 2011.
245
Simon, Herbert: The Sciences of the Artificial (Second Edition) The MIT Press, Cambridge, Massachusetts, 1981 pp 132-133.
Sinnott, D.H.: The Development of Over-the-Horizon Radar in Australia, DSTO Bicentennial History Series, Canberra 1988, available at http://www.dsto.defence.gov.au/attachments/The_development_of_over-the-horizon_radar.pdf. accessed 1 December 2013.
Sinnott, D. H.: John Strath, unpublished monograph.
Smith, Ron: Military Economics – The Interaction of Power and Money, Palgrave, New York, 2009.
Stoneman, P. (ed.) Handbook of the Economics of Innovation and Technological Change, Blackwell Publishers, Oxford, 1995.
Swedish Institute Fact Sheet: The Swedish System of Government, page 4, available at http://www.sweden.se/upload/Sweden_se/english/factsheets/SI/SI_FS55z_The_Swedish_System_o f_Government/The_Swedish_System_of_Government_FS55z.pdf, accessed 1 February 2010.
Sykes, T.: The anatomy of a $50m loss, Bulletin, 19 January, 1992.
Taylor, W.J.: Sweden in Taylor, W. And P. Cole: Nordic Defense: Comparative Decision Making, Lexington Books, Lexington, 1985.
The Technical Cooperation Program: Overview, available at http://www.acq.osd.mil/ttcp/overview/ accessed 2 December 2013.
Thomason, J.: Development of Over-the-Horizon Radar in the United States, in Proceedings of the International Radar Conference, Institute of Electrical and Electronic Engineers Conference Publications, 2003.
Tornatzky, Louis G. And Mitchell Fleischer: The Process of Technological Innovation, Lexington Books, Lexington, 1990.
Van Warden, F.: Crisis, Corporatism and Continuity in Grant, W., J. Nekkers and F. Van Waarden (eds): Organising Business for War: Corporatist Economic Organisation during the Second World War, Berg, Oxford 1991.
Von Hippel, Eric: The Sources of Innovation, Oxford University Press, Oxford, 1988.
Walsham, G.: Interpretative Studies in IS research: nature and method, European Journal of Information Systems, Volume 4, 1995, pp74-81.
Walters, Patrick: RAAF in war with Navy on Spending, Sydney Morning Herald Newspaper, 11 March 1987.
Washington Staff Handbook and its successor, the Washington-based Military Multifora Staff Handbook, available at http://climateviewer.com/assets/img/wiretapped/2012-Multifora- Handbook.pdf. accessed 2 December 2013.
246
Wehner, Joachim: Budget Reform and legislative control in Sweden, Journal of European Politics, Vol 14, no 2, 2007, pp313-332.
Westney, Eleanor D. Imitation and Innovation: The Transfer of Western Organisational Patterns to Meiji Japan, Havard University Press, Cambridge, Mass, 1987.
Yin, Robert K.: Case Study Research Design and Methods (Fourth Edition), Sage Publications, Thousand Oaks California, 2009.
Young, Peter: A hard look at Jindalee, Australian, 23 June 1986, p.
Young, Thomas-Durrell: Cooperative Diffusion Through Cultural Similarity: The Post-war Anglo-Saxon Experience in Goldman, E.O. and L. Eliason (eds): The Diffusion of Military Technology and Ideas, Stanford University Press, Stanford. 2003.
Zarzecki: Arms Diffusion: The Spread of Military Innovation in the International System, Routledge, New York,2002.
Zimmerman, David: Britain’s Shield: Radar and the defeat of the Luftwaffe, Sutton Publishing, Stroud, UK, 2001.
Zimmerman, David: Top Secret Exchange: The Tizard Mission and the Scientific War, McGill-Queen’s Press, Montreal and Kingston, 1996.
247
Interviews
2005 Paul Johnson (CEO Lockheed Martin Australia) on 25 July 2005 at Lockheed Martin Australia, Canberra.
2007
The Honourable Kim C. Beazley (Minister for Defence 1984-1990), on 11 September 2007 at Parliament House, in Canberra.
Interview of Mr John Strath (Head Electronic Systems Division, DSTO until his retirement in September 1983) on 21 November 2007 in Adelaide.
2009 Lennart Kallquist (Senior Vice President Corporate Strategy and Business Development, SAAB AB) on 14 November 2009, Stockholm.
Sven Larsson (Product Manager AEW Systems, Saab Surveillance Systems), Carl-Gilbert Lonroth (FMV project manager for Erieye development and procurement, then Vice President, SGS Program, marketing Erieye to Australia), Bengt Isacsson (Director Strategy, Product Management and Marketing, Saab Microwave Systems) on Monday 16 November 2009 at SAAB Head Office, World Trade Centre, in Stockholm.
COL Mats Olofsson (Chief Scientist, Supreme Commander’s Staff (Development), Headquarters Swedish Armed Forces)on Tuesday 17 November 2009 at SAAB Head Office: World Trade Centre, in Stockholm.
Mr Bo Tarras-Wahlberg (Director of Research Strategy and Markets, FOI ) on 17 November 2009 at the SAAB Head Office, World Trade Centre, Kungsgatan, Stockholm.
Joakim Andersson, FSR 890 Program Director, on 18 November 2009 at SAAB Facility, Linkoping.
Commanding Officer and FSR 890 Operations personnel, SwAF F7 Special Wing, on 18 November 2009 at Linkoping.
248
Lars Karlen (Vice President, Market and Business Development, Saab Surveillance Systems); Roland Karlsson (ex Ericsson microwave) on Friday 20 November 2009 at Saab Microwave Systems, Molndahl.
Professor Charles Edquist, 23 November 2009, Lund University, Lund.
2010 Ms Aase Jakobsson on Friday 15 January 2010 at Building F-G-131, DSTO, Fairbairn.
Carl-Gilbert Lonroth (FMV project manager for Erieye development and procurement) on Monday 5 July 2010, Stockholm.
Prof Stefan Axberg (Chair in Military Technology, Swedish National Defence College), on Tuesday 6 July 2010, Stockholm.
Mr Nils Gylden on Thursday 15 July 2010, Stockholm.
249
Parliamentary Hearings The Joint Committee of Public Accounts held a series of hearings as part of its Review of the Jindalee Operational Radar Network. As part of the research for this thesis, the official Hansard reports of the following hearings were reviewed:
1. Tuesday, 23 July 1996, Canberra;
2. Friday 29 November 1996, Canberra;
3. Thursday 5 December 1996, Canberra;
4. Friday 6 December 1996, Canberra;
5. Monday 3 March 1997, Canberra;
Statements to the JORN hearings held by the Joint Committee of Public Accounts and Audit (JCPAA) by the following witnesses are cited in the thesis:
1. Ayers, T.: Testimony to JCPAA Hearings into JORN, Hansard, Canberra, 6 December 1996, page 82.
2. Bardo, W.: Testimony to JCPAA Hearings into JORN, Hansard, Canberra, Friday 29 November 1996, pp4-5.
3. Brennan, M.: Testimony to JCPAA Hearings into JORN, Hansard, Canberra, Thursday 6 December 1996, especially pp 38-39.
4. Gilligan, M.: Testimony to JCPAA Hearings into JORN, Hansard, Canberra, Thursday 5 December 1996, pp24-34.
5. Jones, G.: Testimony to JCPAA Hearings into JORN, Hansard, Canberra, 23 July 1996, page 19.
250
Thesis-Related Publications and Presentations
Publications R. Wylie, S. Markowski and P. Hall: Big science, small country and the challenge of defence system development: An Australian case study in Defence and Peace Economics, Vol 17, No 3, 2006, pp257-272. (peer reviewed article)
R. Wylie, and S. Markowski: Managing the defence value adding chain: Australian procurement of over-the-horizon radar in S. Markowski, P. Hall and R. Wylie: Defence Procurement and Industry Policy: A Small country perspective, Routledge, London, 2010, pp354-370. (book chapter)
P. Hall and R. Wylie: Isolation and technological innovation in Journal of Evolutionary Economics, vol 24, 2014, pp357-376; (peer reviewed article)
Presentations R. Wylie: Defence Innovation in Sweden and Australia: A sectoral system of innovation perspective, paper presented to Defence Innovation Workshop, Swedish National Defence College, Stockholm, 12-13 April 2011.
P. Hall and R. Wylie: Speciation in Technological Evolution: The effect of differential selection environments on a military technology, paper presented to 14th International Schumpeter Society Conference, Brisbane, 2-5 July 2012.
251