Professional Documents
Culture Documents
Research Policy: Collaborative, Pre-Competitive R & D and The Firm
Research Policy: Collaborative, Pre-Competitive R & D and The Firm
Research Policy: Collaborative, Pre-Competitive R & D and The Firm
policy
ELSEVIER
Abstract
Governments in many countries have supported programmes of collaborative, pre-competitive R & D in enabling
or generic technologies. The paper is based on 6 years of research into the UK Alvey advanced information
technology (IT) programme, augmented by research on the Swedish IT4 programme. Our focus is the relationship
between pre-competitive, collaborative R & D and wider firm strategies and processes. The underlying premises of
the paper are that the benefits of pre-competitive R & D are long-term rather than short-term, systemic rather than
discrete, and accessible only through analysis of processes rather than by counting outputs or calculating return on
investment. The research provides insights into the strategies and mechanisms of firms' participation, modes of
collaboration, and the processes of technology transfer, knowledge assimilation and application within a firm.
1. Introduction
Industrial collaboration in research and development ( R & D ) has been a major feature of
"The paper is based on research funded by the UK Department of Trade and Industry and the Science and Engineering Research Council, undertaken as part of the SPRU
evaluation of the Alvey Programme. Paul Quintas gratefully
acknowledges the subsequent support of the UK Economic
and Social Research Council's Programme on Information
and Communication Technologies. Additional material draws
on the evaluation undertaken by Technopolis Ltd., funded by
the Swedish Ministry of Industry, of the Swedish IT4 programme. The authors thank SPRU colleagues Dr. Mike Hobday, Dr. Mark Dodgson, Dr. Robin Mansell and Professor
Keith Pavitt for comments on a previous draft, and the
anonymous US reviewers for their valuable guidance. Responsibility for the final version remains, of course with ourselves.
Formerly at: Science Policy Research Unit, Mantell Building,
University of Sussex, Falmer, Brighton, UK.
* Corresponding author.
326
3 The pragmatic if atheoretical working definition of 'precompetitive' R&D that emerged in the 1980s was 'that R&D
which competing firms will agree to collaborate on'.
ration, technology transfer and the internal corporate processes of technology acquisition, learning and assimilation. 4 It is necessary, however,
briefly to highlight the central concerns underpinning the questions and hypotheses which form
the basis of the empirical enquiry.
The relationship between public policy and
corporate technology strategy is of primary interest. It has long been accepted that, unaided, the
market invests sub-optimally in research [7,32,39].
Through the 1980s, even the UK Conservative
government under Margaret Thatcher (extreme
espousers of the sovereignty of the market) continued to acknowledge that market mechanisms
fail to ensure that adequate resources are allocated to R & D [14]. The case for the Alvey Programme was largely formulated in terms of market failure arguments, as we describe below.
However, political considerations imposed constraints on the type of intervention sanctioned. In
particular, and in addition to the focus on precompetitive R & D, the UK government sought to
protect the interest of the tax-payer by ensuring
that firms participating in such R & D programmes use public funding for additional work;
that is, for R & D which would not have occurred
in the absence of government support. This requirement for 'additionality' clearly raises questions about the relevance of such research to
firms' broader technology strategies and business
objectives, an issue returned to in Section 4.
Alvey provides an opportunity to analyse R & D
collaboration at the level of the participant firms.
Such an inquiry must be located within the
framework of an understanding of the function of
R&D, such as that provided by Freeman [18]. In
particular, we know that firms, in addition to the
pursuit of wholly internal projects, undertake R&
D in order to be able to track external develop-
4 A comprehensive introduction to many of the issues relating to inter-firm cooperation may be found In Contractor and
Lorange [13]. See also Mowery [29,30]. Badaracco [8] contains
insights into competitiveness through collaboration. Watkins
[56] presents a concise review of the literature relating to
R&D cooperation. Technological learning within organisations is addressed in Argyris and Schon [5] and Fiol and Lyles
[16]. For a review of IT R&D policy see Arnold and Guy [6].
327
328
Throughout the 1980s, government and industry in many countries supported programmes of
collaborative R & D in IT, involving firms, research institutes and academic institutions [6].
Examples include the US Microelectronics and
Computer Technology Corporation (MCC), the
US Semiconductor Manufacturing Project (Sematech), the Japanese TRON, Sigma, and Fifth
Generation Programmes, the European Community's ESPRIT, RACE and other programmes,
and the pan-European Eureka Programme [6]. In
this paper, the focus is on the UK Alvey Programme, which ran from 1983 to 1989. 7
Alvey was a national programme launched by
the UK government largely in response to the
Japanese Fifth Generation Computer Programme
announced in 1981. The Japanese strategy galvanised action in many countries by threatening a
quantum leap to next-generation computing. In
the UK the announcement coincided with concern over the competitive failure of much of the
IT supply industry. The UK IT sector had been in
relative decline for over a decade, performing
especially poorly in the computer, consumer elec-
329
330
base for further R & D, particularly within a European collaborative context. In a departure from
the pre-competitive focus, Alvey also supported
five Large Scale Demonstrator projects which
were intended to provide exploitation-led goals
for Alvey R & D projects, and to demonstrate the
results of Alvey within the programme's lifetime.
Some dissemination and awareness activity was
also pursued on a small scale, the most successful
mechanisms being technology-specific clubs which
provided fora for interaction.
Government support for R & D collaboration
between firms raises a number of issues relating
to corporate R & D strategy. The following section discusses the interaction between government policy and R & D strategy at the level of the
participant firm.
331
332
firms will be prepared to undertake research projects which on strict commercial criteria they do
not themselves consider worthwhile. Such projects are therefore marginal, as argued by IT
industrialist Rob Wilmot (former chief executive
of ICL), who states that there is no point in
governments "bringing marginal projects into life
with subsidies" precisely because they are, by
definition, not central to the firms' core activities
[59]. This argument is in direct opposition to
government's desire for R & D to be 'additional'.
The distinction between core and marginal
R&D, and the related 'additionality' question,
thus require discussion. Pavitt et al. [41] show
empirically that in most UK industrial sectors,
firms' innovative activity is concentrated within
their principal sector of competence. The authors
conclude that technological knowledge is cumulative and differentiated, and that "firms' technological strategies are heavily constrained by their
core activities, their size, and their forms of organisation" [41, p. 64]. However, concentration
on current core technologies may make the firm
vulnerable to threats from new technologies (e.g.
semiconductors for 1950s vacuum tube makers,
laser technology for 1970s photocopier manufacturers). Technology regarded as being peripheral
Number of Respons
40
30
20
10
C
Long Term
Core Technology
Peripheral Technology
. . . . . . . . . .
II
333
Number of Respon:
30
20
10
Long T e r m
Core Technology
..........
Peripheral Technology
;i
technologies, which are the current central technological concerns of firms, and peripheral technologies, here defined as possible replacement
technologies, i.e. alternative technological trajectories. The right-hand scale shows a spectrum of
work running from R & D with very short-term
objectives to work which has much longer-term
horizons. Participant firms were asked to characterise their Alvey projects along these two dimensions using 1-5 scales as indicated.
With firms characterising their own in-house
R & D in this fashion, we might expect the bulk of
R & D projects to lie in the core technology/
short-term sector (i.e. the left-hand corner), with
some of the richer, larger a n d / o r more innovative firms supporting longer-term work in core
areas. In contrast to expectations, fig. 1 demonstrates that Alvey allowed firms to expand their
R &D portfolios in two major ways. In the first
instance, the programme allowed firms to conduct more long-term R & D in their core areas, as
illustrated by the cluster at the back of the matrix. This is the kind of work that is on the 'wish
list' of firms; usually it is only financial or other
resource constraints which limit a firm's ability to
undertake this work. It is mistaken to think of
such R & D as unimportant or marginal; rather, it
is more fitting to think of it as luxury R & D ,
334
nition, since the technologies explored are peripheral). However, this is exactly the kind of
work which is facilitated by collaborative R & D .
Collaboration allows cost-sharing, leverage and
access to complementary expertise that is too
costly and risky to acquire (because of its peripheral status) via go-it-alone routes. Alvey allowed
firms to hedge their technological bets by facilitating the pursuit of insurance R & D . Moreover,
fig. 2 indicates that Alvey was not alone in doing
this. Work undertaken in the context of the
Swedish IT4 p r o g r a m m e - - a programme closely
resembling Alvey, running initially from 1987 to
1990--supported an almost identical pattern of
work.
We can look in more depth at the ways in
which Alvey participation brought benefits to the
companies involved. Cross-checking between a
range of questions put to participants verifies the
previous findings concerning peripheral and
long-term R & D . Table 1 shows that 45% of
industrial teams said that Alvey was very important in enabling them to enter new R & D areas,
and 46% said that it had enabled them to access
academic knowledge. Twenty-eight per cent of
industrial participants stated that Alvey had been
very important in enabling them to build critical
mass in an R & D area (i.e. where capability was
previously sub-critical mass), and 27% stated that
their participation had enabled them to keep
track of R & D developments. More directly relevant to core activities, Alvey allowed firms to
accelerate R & D activity beyond the pace that
would otherwise have been achieved (57% of
teams), to 'maintain an R & D presence' (46%)
and 'build on their R & D base' (45%). Similarly,
Alvey enabled 35% of firms to deepen their understanding in a particular technology area. These
benefits--accelerated R & D , maintained R & D
and deepened knowledge--suggest that firms
used the programme to enhance and support
pre-existing technological strategies in addition to
entering new R & D areas.
We also need to take account of the fact that
firms' business and technology strategies may
change radically over a period of 5 years, and
firms may decide to pull out of a business area,
e.g. STC's decision to withdraw from the mer-
Table 1
What Alveyenabled industrial participants to do
Activity
% of teams Achievedeclearing ment
score b
Alvey to
be very
important a
Develop new tools and techniques
Accelerate R&D
Upgrade skills
Use new tools and techniques
Maintain R&D presence
Access academic know-how
Enter new R&D areas
Build on R&D base
Enhance image/reputation
Establish new academic links
Spread costs
Deepen understanding
Develop prototypes
Achieve critical mass in area
Keep track of R&D
Spread risks
Develop products
Establish new industry links
Access industry know-how
Enter international R&D
programmes
Enter private sector R&D ventures
Enter national R&D programmes
Get used to new IT standards
Influence new IT standards
Enter new non-R&Dcollaborations
64
57
52
51
46
46
45
45
38
38
37
35
35
28
27
25
22
22
22
22
3.7
3.8
3.9
3.9
3.6
3.2
3.8
3.8
3.7
3.5
3.6
3.8
3.5
3.6
3.6
3.8
3.0
3.1
3.1
3.3
20
15
10
10
6
3.2
3.2
3.2
3.1
3.1
chant semiconductor business, or GEC's withdrawal from the CMOS semiconductor market,
both of which happened during the Alvey Programme. Therefore, it should not be assumed
that core technology is immutable and thus that
investment in core R & D is devoid of risk.
One of the most significant result in table 1 is
that 46% of industrial participants stated that
Alvey had enabled them to maintain a presence
in an area which would otherwise have been cut.
There were also examples of firms admitting that
they could not have undertaken the R & D necessary for their core business activities (e.g. C A D
R & D in the case of a semiconductor producer)
without the provision of Alvey funding, because
o,,
335
different technology areas, to the structural position and strategic orientation of R & D within
companies, and to the size of projects--large
commitments requiring higher-level decisionmaking. In particular, the central core of the
VLSI programme required a 'top down' approach
by participant firms--a coherent set of projects
clustered around whole-process projects, as shown
in fig. 3. This represents the portfolio of VLSI
projects of one firm, showing how six smaller
projects feed into the large whole process project
over time. This coherent approach contrasts with
firms that had numerous isolated Alvey participations, each driven by one or two researchers with
little wider support within the firm. Coherent
strategies were not restricted to the VLSI area.
ICL's involvement in the Systems Architectures
sub-programme, which had significantly more distant horizons than VLSI R & D, featured a cluster
of related projects, including several long-term
research projects addressing alternative technological paths.
A second way of looking at firms' Alvey strategies is to focus on structural and organisational
factors within the firm, in particular, the degree
of integration or autonomy of the R & D department undertaking the Alvey work. One indication
of R & D departmental autonomy is control over
funding decisions. Alvey project budgets were
approved by the firms' R & D management (as
making project
etchn.,ojc,k
Lithography
project~\
equipment
Measurin9 equipmentproject
projects
1984
1985
1986
1987
1988
Fig. 3. Example of coherent firm strategy. Note the seven horizontal lines each represent one Alvey project. Results from the
Techniques projects and Equipment projects feed into the Whole Process project over time, as shown by the angled lines.
336
Table 2
R & D outputs: expectations and achievements a
Expected outputs
Percentage
of teams
expecting
outputs ~
Hardware
Software
Neither
17
68
27
3.7
3.6
-
18
76
21
3.6
3.9
-
Research outputs
Test and m e a s u r e m e n t
Specifications/designs
Evaluations
Feasibility statements
Simulations
Standards recommendations
Demonstrators
Total research outputs d
12
38
41
26
16
1l
36
84
3.7
3.9
3.7
3.8
4.2
3,8
3,7
-
Development outputs
Prototypes
Marketable products
Neither
Total development outputs d
36
14
8
43
3,8
3,2
-
Mean
output
achievement
scores
Table 3
R & D financial autonomy and R & D performance a
Project budget
approved by
Satisfactory
Higher m a n a g e m e n t
R&D management
27
7
29
37
56
44
Total
34
66
100
Total
337
Table 5
Links to other Alvey projects and R & D performance a
Table 4
R & D locational separation and R & D performance a
Location of R & D
and production
Satisfactory
23
9
27
41
50
50
Total
32
68
100
Total
Rating of project
performance b
Unsatisfactory
Satisfactory
Total
47
17
64
15
21
36
Total
62
38
100
a Data are for projects aimed at 'development-oriented' obj e c t i v e s - p r o t o t y p e s and marketable products.
b Percentage of industrial teams rating their project output
performance satisfactory or unsatisfactory.
Statistical significance: X 2 = 3.5; significance = 0.062.
Source: S P R U Final Alvey Questionnaire [21].
Table 6
In-house R & D links and Alvey R & D performance a
Links with other R & D
in the firm
Rating of project
performance b
Unsatisfactory
Satisfactory
Total
26
17
43
18
38
56
Total
44
55
100
a Data are for projects aimed at 'development-oriented' obj e c t i v e s - p r o t o t y p e s and marketable products.
b Percentage of industrial teams rating their project output
performance satisfactory or unsatisfactory.
Statistical significance: X 2 = 3.5; significance = 0.062.
Source: S P R U Final Alvey Questionnaire [21].
338
tion to these data, success in development-oriented projects was also found to correlate strongly
with high levels of project effort in terms of
person/years.
In contrast to these results concerning development-oriented projects, those projects geared
to research-oriented objectives tended to achieve
satisfactory performance where the Alvey R& D
projects had no connections to other Alvey projects conducted by the firm (i.e. no coherent
portfolio), or connections with other in-house R
&D. In other words, research-oriented projects
had greater success where these were individual,
isolated efforts, whilst development-oriented projects benefited from linkage to other R & D in the
firm, both Alvey and non-Alvey.
This complex picture points to some tentative
conclusions, although these are strongly supported by qualitative interview material. Coherent strategies, where Alvey projects were part of
a portfolio or linked to in-house R & D , were
correlated with successful achievements in the
case of projects aimed at the development end of
the R & D spectrum. Coherent strategies were
not, however, correlated with success in the case
of research-oriented projects. Here, isolated projects achieved satisfactory outcomes, and success
was strongly correlated with the relative autonomy of R & D units,
Further analysis reveals other correlations between R & D organisation and strategy and satisfactory project achievements. In the case of projects geared to software outputs, success was
linked to R & D autonomy, both in terms of budget control and location. Success in these cases
was also linked to projects being considered core
R&D: that is, R & D perceived as highly necessary, often with a high level of effort committed
in terms of person/years.
Success in project terms does not, however,
indicate that the R & D results have been taken
up within the wider firm context. This points to
the need to look at 'internal' technology transfer
and assimilation within participant firms. Before
turning to this, however, we look at participants'
experience of Alvey R & D collaboration in order
to understand why and how collaboration occurred.
339
340
Table 7
Follow-up R&D of Alvey industrial participants
Percentage of teams in each technology area
with each follow-up action
No follow-up
R&D
One firm only
With Alvey
academic
partners
With Alvey
industrial
partners
With new
partners
Team numbers
in sample
SE
IKBS
VLSI
MMI
LD
Total
28
17
19
19
31
22
35
12
45
27
30
18
33
26
19
38
36
29
11
11
15
27
13
22
41
25
35
45
30
18
27
32
26
11
114
Table 8
Transfers from Alvey industrial teams
Recipient
environment
34
22
43
15
Output
in
regular
use
29
17
341
342
Table 9
Technology area variation in transfers
Recipient
environment
Output
evaluated
Output
further
developed
Output in
regular
use
VLSI
SE
IKBS
MMI
40
38
27
56
VLSI
SE
IKBS
MMI
33
24
23
44
VLSI
SE
IKBS
MMI
50
43
50
31
VLSI
SE
IKBS
MMI
53
14
19
22
VLSI
SE
IKBS
MMI
25
33
23
22
VLSI
SE
IKBS
MMI
25
33
15
16
VLSI
SE
IKBS
MMI
10
19
19
16
VLSI
SE
IKBS
MMI
25
14
15
16
Figures are the percentage of industrial teams in each technology area indicating each type of transfer.
a Transfer categories are not mutually exclusive.
VLSI, Very Large Scale Integration; SE, Software Engineering; IKBS, Intelligent Knowledge Based Systems; MMI; Man-Machine
Interface.
Source: SPRU Final Alvey Questionnaire [21].
343
9 As part of the Programme for Information and Communication Technology (PICT) funded by the U K Economic and
Social Research Council, Paul Quintas is following up some
Alvey Software Engineering projects in order to track postprogramme technology transfer and use.
344
8. Conclusions
Drawing on the experience of the UK Alvey
programme, this paper provides an analysis of the
conduct of collaborative, pre-competitive R & D
at the level of the participant firm. Our purpose
has been to further current understanding of the
strategies and processes which frame such R & D
within companies.
We have differentiated a number of strategies
employed by firms in their approach to Alvey
R&D, and the analysis has linked strategy with
R & D outcomes. In evaluating the benefits of
programmes such as Alvey, the added value is not
simply measured in terms of research undertaken
that would not otherwise have been funded; the
benefits are also the acceleration, broadening and
deepening of R&D, and the maintenance of R&
D which may have been cut. Alvey also enabled
firms to undertake longer-term R & D in core
technology areas, and increased their ability to
carry out insurance R & D in peripheral areas.
Alvey projects covered a wide range, from
'unspectacular engineering' aimed at the next
incremental step straight down the technological
highway, to ambitious objectives, some located on
the far horizon along the main route, and others
out in uncharted territory. There are significant
differences between the policies and R & D
strategies appropriate for these. For those projects aimed at the development end of the R & D
spectrum, a coherent strategy in which Alvey
projects were clustered into a tight portfolio, or
linked to in-house R&D, was correlated with
successful achievements. In the case of researchoriented projects (i.e. those focused on enabling
technology), analysis linking strategy with organi-
sational structure revealed that success was correlated with the relative autonomy of R & D units.
Isolated projects achieved the most satisfactory
outcomes for researchers. This suggests that different management strategies and policy mechanisms are required according to the focus of the
R&D.
Alvey's encouragement of collaboration succeeded in uniting and focusing formerly fragmented resources, and building networks of researchers. Many project teams have learned how
to collaborate, and gone on to further R & D
alliances, particularly within European programmes. As a mechanism to enable participants
to access technology and absorb know-how from
their partners Alvey was a qualified success, with
much variation between firms, collaboration types
and (in particular) between technology areas. Almost half of the industrial participants stated that
Alvey collaboration had been very important in
enabling them to access technology and absorb
know-how from their academic partners, compared with 22% when the source was a collaborating industrial organisation. Success in accessing knowledge and technology through collaborative R & D is a function of many factors, including
the compatibility of collaborative partners, the
mode of collaboration, whether codified or tacit
knowledge is being transferred, the relative maturity of the technology area, and stability within
both supplier and participant organisations. We
have seen that Alvey collaborations took several
forms, and it is clear that different arrangements
are appropriate for different project types.
It should be noted that responses to direct
questions concerning knowledge and technology
transfer inevitably focus on codifiable knowledge,
tending to underestimate the extent to which
personal contact and collaborative working leads
to transfers of tacit knowledge. Interviews both
during and after the programme suggest that
Alvey did support tacit knowledge transfers, but
the extent of this is impossible to estimate. One
partial indicator is personnel movement. An outflow of 18% of research staff from R & D units
undertaking Alvey projects to external firms indicates a potential for tacit knowledge flows between organisations. Subsequent data suggest that
345
346
market, GEC's withdrawal from CMOS semiconductor technology, and the volatile conditions
generated by the series of takeovers and mergers
in the IT sector. Counter to the Sciberras argument raised in Section 4, STC, GEC, Plessey and
Ferranti's VLSI strategies could hardly have been
labelled 'marginal' R & D in 1984, but they were
overtaken by business-driven change which, in
some cases, negated the investment. This again
suggests that a focus on enabling technologies
which maximise future flexibility is more appropriate than targeting government investment towards current core business areas.
Conversely, the success of ICL's Flagship project in leading (over 10 years and through additional collaborative R & D within ESPRIT) to the
Goldrush parallel computer illustrates the value
of supporting long-term collaborative R & D in
core technology areas. However, much of the
R & D undertaken in Flagship and the cluster of
projects surrounding it was pursuing enabling,
generic technologies (e.g. declarative systems) and
alternative paths leading into peripheral technological areas. Not all of these paths were ultimately followed, but their pursuit was necessary
because the appropriate technological choices
could not be predicted with certainty in 1984.
This raises a more general question concerning
the limitations of present-day assessment of the
benefits of pre-competitive R & D in enabling
technologies. Whilst many projects reported at
the end of Alvey that their company was not
interested in further investment in the R & D
project and that 'exploitation' was not happening,
this is not necessarily the end of the story, as the
recent research described in the previous section
has revealed. This indicates that the 'exploitation'
of pre-competitive R & D may be a non-linear
process involving serendipitous links and unpredictable outcomes. Therefore, assessment of the
benefits of Alvey cannot easily be bounded in
time or scope.
Whilst the particular circumstances pertaining
in the UK, such as the investment climate and
prevailing business ethos, may not directly parallel experience elsewhere, the central lessons of
Alvey are nevertheless of relevance to other
countries. Our findings and recommendations
9. References
[1] T.J. Allen, Managing the Flow of Technology (MIT Press,
Cambridge, MA, 1977).
[2] Alvey Committee, A Programme for Advanced Information Technology (Department of Industry, HMSO, London, 1982).
[3] Alvey Directorate, Alvey Programme Annual Report 1987
(Department of Trade and Industry, London, 1987).
[4] M. Aoki, Horizontal vs. Vertical Information Structure of
the Firm, The American Economic Review 76 (5) (1986)
971-983.
[5] C. Argyris and D.A. Schon, Organizational Learning: A
Theory of Action Perspective (Addison-Wesley, Reading,
MA, 1978).
[6] E. Arnold and K. Guy, Parallel Convergence: National
Strategies in Information Technology (Frances Pinter,
London, 1986).
[7] K.J. Arrow, Economic Welfare and the Allocation of
Resources for Invention, in: R. Nelson (Editor), The Rate
and Direction of Inventive Activity (Princeton, University
Press, Princeton NJ, 1962).
[8] J.L. Badaracco, The Knowledge Link: How Firms Compete Through StrategicAlliances (Harvard Business School
Press, Cambridge, MA, 1991).
[9] R.A. Burgelman and R.S. Rosenbloom, Technology
Strategy: An Evolutionary Process Perspective, Research
on Technological Innovation, Management and Policy, Vol.
4 (JAI Press, Greenwich, CT, 1988).
[10] T. Burns and G.M. Stalker, The Management oflnnnovation (Tavistock, London, 1961).
[.11] W.M. Cohen and D.A. Levinthal, Innovation and Learning: The Two Faces of R&D--Implications for the Analysis of R&D Investment, The Economic Journal 99 (1989)
569-596.
[12] Computing, ICL Strikes Gold in Rush to Claim Parallel
Market, Computing 28 October (1993) 8.
[13] F. Contractor and P. Lorange, Cooperative Strategies in
International Business (Lexington Books, Lexington, MA,
1988).
[14] Department of Trade and Industry, DTI: The Department of Enterprise (HMSO EM278, London, 1988).
347
348