Absorptive Capacity and Regional Patterns of InnovationMaria Abreu, Vadim Grinevich,
Michael Kitson and Maria Savona
Presentation Structure• Regional Growth in the UK• The CBR Survey: A Meta-Regional Picture• The CBR Survey: The Innovation Picture• The Absorptive Capacity Concept• Innovation and Absorptive Capacity: Multivariate
Analysis• Innovation and Absorptive Capacity: ‘Open’ v
‘Closed’ innovation• The Case Study Evidence• Implications
Regional Productivity Disparities: The Scale of the Problem
• Since the mid-1980s, major productivity gap has opened up between London and South East on one hand, and rest of UK on the other
• Only three regions have productivity above UK average (London, South East and Eastern)
• Productivity growth in the Northern regions has been especially slower
• The productivity gap is such that output per employee in South East now more than 28 percent above that in Northern Ireland, and 20 percent above that in North East
GVA per capita, 2003
Source: Local Knowledge 0 550,000 1,100,000275,000 Meters
GVA per capita (£)
11,000 - 12,000
12,000 - 13,000
13,000 - 14,000
14,000 - 15,000
15,000 - 40,000
Change in GVA per capita, 1995-2003
Source: Local Knowledge 0 550,000 1,100,000275,000 Meters
Change in GVA per capita (%)
40 - 42
42 - 44
44 - 46
46 - 48
48 - 50
50 - 60
Regional Productivity Disparities: The Scale of the Problem
• Employment growth has also been unequal across the regions
• Highest in South East, Northern Ireland, South West and Eastern region
• Some regions have not fully recovered from massive job declines of early-1980s (North East, North West and West Midlands)
• South East only region to experience above average growth in productivity and employment
Drivers of Regional Competitiveness/Productivity According to HM Treasury/DTI
SKILLS ENTERPRISE
INNOVATION COMPETITION
INVESTMENT
Entry of newfirms raisescompetition
Skills raise firms’capacity to
develop and usenew technology
Management skills raiseentrepreneurship andbusiness excellence.New firms createdemand for skilled
labour
Increasingcompetitionencouragescompetition
Increasingcompetition creates
incentives forbusiness investment
Investment inphysical capitalincreases firms’
innovative capacity
Are innovative regions, prosperous regions?
Source: Martin (2005)
CBR Dataset
• Centre for Business Research (CBR) Survey 2004• 1-499 employees• Manufacturing and business services• 2137 respondents • More detailed information than in CIS although the
CIS survey is larger • CBR Survey also allows the opportunity to review
and research respondents in greater depth
Regional Competitiveness
• For the purpose of this initial analysis, a meta-regional classification is used: – Greater South East (South East, Eastern Region
and London)– Middle England (East Midlands, West Midlands
and the South West)– The Northern Way (North West, North East and
Yorkshire and Humberside) – Scotland/Wales
Markets and Competition
• The Greater South East engaged in the most international trade with over 20% of the businesses identifying overseas markets as being most important compared to only 12% of firms in Scotland/Wales
• There was a more intense competitive environment in the Greater South East with evidence of a core to periphery decline in the intensity of competition
• The Greater South East had the highest level of international competition – with 41% of firms facing some overseas competitors and with 10% of firms competing exclusively with overseas businesses
Geographical Distribution of Markets
0
10
20
30
40
50
60
Greater SouthEast
MiddleEngland
Northern Way Scotland/Wales
% D
istr
ibu
tio
n o
f F
irm
s
Local
Regional
National
International
Number of Serious Competitors(% distribution of firms)
Number Of firms
Greater South East
Middle England
Northern Way
Scotland & Wales
0 10.2 11.9 12.3 14.3 1 – 4 34.5 34.3 35.7 43.5 5 – 9 24.3 23.5 28.5 19 10 – 19 15.6 16 13.9 16.7 20 – 98 11.5 10.4 6.7 4.8 99 + 3.9 3.9 2.8 1.8 Total Responses 826 463 389 168
Extent of Overseas Competition(% distribution of firms)
% Greater
South East Middle
England Northern Way Scotland &
Wales 0 59.5 60.2 65.1 67.6 0 - 24 7 8.6 6.9 5.1 25 - 49 9.1 7.6 8.2 9.6 50 - 74 10.5 10.2 11.6 7.4 75 - 99 4.2 3.9 4.1 6.6 100 9.6 9.4 4.1 3.7 Total Responses 684 382 318 136
Networks and Cooperation
• The use of networks and effective collaboration has been identified as an important factor in improving business and regional competitiveness
• The highest level of collaboration took place in the Greater South East (48% of businesses) and the lowest took place in the Northern Way (37% of businesses)
Use of Collaboration and Networks
0
10
20
30
40
50
60
GreaterSouthEast
MiddleEngland
NorthernWay
Scotland/Wales
% o
f F
irm
s
% of Firms
The Geography of Networks and Cooperation
• There were significant regional variations in collaborative partners – illustrating that there are significant differences in the character
of, and scope of, regional networks • Firms in similar lines of businesses were the most important
collaborative partners– but this group was more important for Scotland/Wales and the
Greater South East compared to Middle England and the Northern Way
• Forward and backward linkages (collaboration with customers and suppliers) were more important for businesses in Middle England
• Collaboration with universities and research institutes was higher in the Northern Way compared to other regions
Policy Support
• The use of Business Link (and Business Gateway and Business Connect) was significantly greater in Middle England and the Northern Way than in the Greater South East and especially compared to the low use in Wales/Scotland
• The use of Local Learning and Skills Councils was higher in Middle England and the Northern Way compared to the other two regions
• The assistance from Development Agencies had a distinct core periphery incline – with the lowest use in the Greater South East and the highest (by far) in Scotland/Wales
• There was core periphery incline for export advice- with the highest level in Scotland/Wales
Policy Support
Regional Variations in use of Government sponsored business support agencies
(% distribution of firms)
Greater South East
Middle England
Northern Way
Scotland & Wales
Business Link/ Gateway/Connect*** 31.7 44.3 46.3 17.9 Local Enterprise Agency*** 5.5 7.6 12.5 28.3 Local Learning Skills Council *** 7.3 12.3 12.9 8.7 RDA/ WDA/ SE*** 4.4 8.0 11.8 37.5 Export Advice 9.6 13.3 12.0 14.7
Innovation Behaviour• Approximately 60% of firms introduced product
innovations • Approximately half of firms had introduced process
innovation• The CBR data indicates a higher level of innovation than
CIS– Different survey structures, CBR covers
manufacturing and business services– Issue of weighting– Issue of questionnaire design
Innovation Performance
• Product and Process innovation– Positively associated with firm size– Positively associated with firm growth– Higher in high-technology sectors
Product innovationIntroduction of product innovations during the previous 3 years
59.4
42.0
65.7
72.5
53.9
64.3
73.2
59.1
74.6
50.9
31.9
21.7
34.8
42.3
27.6
35.7
45.0
30.1
48.7
25.2
0
10
20
30
40
50
60
70
80
% o
f F
irms
Innovators Novel innovators
Source: Fu (2006)
Process innovationIntroduction of process innovations in previous 3 years
49.3
32.3
55.2
62.3
45.4
53.8
58.2
52.255.4
40.7
19.2
12.6
19.7
30.5
17.1
21.3
15.5
19.3
28.5
17.1
0
10
20
30
40
50
60
70
% o
f F
irms
Innovators Novel innovators
Source: Fu (2006)
Region Total Firms % Innovative Firms
% Goods Innovation
% Service Innovation
% Process Innovation
London 290 52.22 24.11 30.11 33.95
Eastern 257 52.87 27.87 32.52 43.32
South East 344 44.44 26.49 22.59 28.50
South West 181 56.88 31.08 30.35 42.40
West Midlands 185 50.72 28.92 19.86 35.40
East Midlands 139 53.09 26.40 18.82 41.45
Yorkshire 177 57.16 29.09 31.77 47.82
North West 203 46.17 25.94 23.90 26.56
North East 51 56.43 36.18 26.27 21.28
Wales 68 46.27 30.67 26.81 27.01
Scotland 111 56.92 35.11 38.87 34.01
Total 2006 52.11 29.31 27.44 34.67
Innovation Statistics by Region, CBR SME Survey 2004
Absorptive Capacity: “…the ability of a firm to recognize the value of new, external information, assimilate it, and apply it to commercial ends is critical to its innovative capabilities.” (Cohen and Levinthal, 1990)
• What are the main determinants of absorptive capacity? • Can regional variations in absorptive capacity explain regional
differences in innovation?• Can we draw implications for regional and national innovation
policies?
Research Questions
Absorptive Capacity
• Originally developed in the cognitive sciences (Bower and Hilgard, 1981; Lindsay and Norman, 1977)
• Extended by Cohen and Levinthal (1989, 1990) to the firm level:– Firm-level absorptive capacity is more that the
sum of the absorptive capacities of individual workers.
– Better internal communication processes improve the firm’s ability to absorb external knowledge.
– Networks are necessary to identify sources of relevant external knowledge.
• Used at the national level to denote the conditions necessary for a technological catch-up to begin (Nelson and Phelps, 1966; Verspagen, 1991; Keller, 1996; Criscuolo and Narula, 2002)
• Extension to the regional level:– Regional absorptive capacity is more than the sum of
the absorptive capacities of individual firms.– Role of local labour markets (Roper and Love, 2006)– Regional networks allow firms to identify potential
sources of knowledge.– Regional development policies promote firms’
investment in absorptive capacity.
Absorptive Capacity
Methodology• Empirical analysis of the effect of absorptive capacity on:
– The probability of introducing a product, service and process innovation.
– The probability of developing an innovation within the firm, in collaboration with other firms or institutions, or by adoption from other firms or institutions.
• Multivariate Probit model to improve efficiency and exploit all available information (Greene, 2003, pp. 174-175).
Data Issues
• Absorptive capacity variables:• Percentage of R&D employees• New management techniques• Employee training
• Connectivity variables• Collaborations with different spatial dimensions
Data Issues
• Control variables: • Exporter• Age, size and size squared• Ownership structure• Market• Number of competitors• Perceived obstacles to innovation• R&D tax credit• Sectoral and regional dummies
Variable Manufactured Product
Service Product
Manufacturing Process
Service Process
R&D Employees 0.449** 0.529* 0.009 0.687**
Managerial Training -0.010 0.520*** 0.091 0.322**
Scientist Training -0.056 -0.222 0.013 -0.289**
Other Employee Training 0.341** -0.078 0.000 -0.027
Total Quality Management
0.154 0.268** 0.322*** 0.145
Quality Circles 0.289* 0.102 0.316** 0.161
Job Rotation 0.282** 0.020 0.400*** 0.123
Performance-Related Pay
-0.033 0.080 -0.061 0.027
Collaborations: Local 0.044 0.123 0.341** 0.213
Collaborations: National 0.209* 0.340*** -0.052 0.287**
Collaborations: Overseas
0.482*** -0.033 0.269** 0.016
MV Probit Estimates, CBR SME Survey 2004
Variable Manufactured Product
Service Product
Manufacturing Process
Service Process
Obstacles: Finance 0.171 -0.022 0.123 0.053
Obstacles: Skills -0.031 0.160 -0.058 0.337**
Obstacles: Technology 0.269** 0.020 0.066 0.195*
Obstacles: Market 0.086 0.089 0.029 -0.195*
Obstacles: Other -0.283** -0.058 -0.252* 0.039
R&D Expenditure / Turn. 0.049 0.003 0.060 0.014
Claimed Tax Credit 0.571*** -0.028 0.199 -0.231*
Business Advice: Bus. Link
0.069 0.081 0.287** 0.141
Business Advice: RDA -0.088 -0.332** -0.072 -0.265*
MV Probit Estimates, CBR SME Survey 2004, cont.
Variable Manufactured Product
Service Product
Manufacturing Process
Service Process
Market: Local reference reference reference reference
Market: Regional 0.230 0.147 -0.141 0.082
Market: National 0.268 0.152 0.105 0.172
Market: International 0.497** 0.100 -0.003 -0.030
Age (Years) -0.004** -0.001 -0.001 0.001
Size: Employees (Ln) 0.032 0.117** 0.110** 0.075
Ownership: Proprietorship
reference reference reference reference
Ownership: Partnership -0.646** -0.604** -1.028*** -0.074
Ownership: Company -0.537* -0.165 -0.595** -0.091
Ownership: Other -4.368 -0.373 -5.166 -4.602
Observations: 805
Log Likelihood -1598.04
MV Provit Estimates, CBR SME Survey 2004, cont.
Absorptive Capacity: R&D
Manufactured Product
Service Product
Manufacturing Process
Service Process
R&D
Expenditure0 0 0 0
R&D
Employees+ + 0 +
Innovation Type
Absorptive Capacity: TrainingManufactured Product
Service Product
Manufacturing Process
Service Process
Managerial
Training0 + 0 +
Scientist
Training0 0 0 -
Absorptive Capacity: Management Practices
Manufactured Product
Service Product
Manufacturing Process
Service Process
Total
Quaity
Management
0 + + 0
Quality
Circles+ 0 + 0
Job
Rotation+ 0 + 0
Connectivity: the Role of CollaborationManufactured Product
Service Product
Manufacturing Process
Service Process
Local 0 0 + 0
National + + 0 +
Overseas + 0 + 0
The Role of MarketsManufactured Product
Service Product
Manufacturing Process
Service Process
Regional 0 0 0 0
National 0 0 0 0
Overseas + 0 0 0
The Impact of PolicyManufactured Product
Service Product
Manufacturing Process
Service Process
R&D Tax Credits
+ 0 + -
Business Advice: Business Link
0 0 + 0
Business Advice: RDA
0 - 0 -
Summary of the CBR results
• Absorptive capacity variables have a positive effect on innovation
• Management practices are particularly important for manufacturing product and process innovation.
• Local collaborations are only important for manufacturing process innovation
• National and overseas collaborations are generally more important than local collaborations
• Sectoral specificities are strong and significant.• Regional dummies are not significant.
Regional Effects
• Regional impacts captured through variations in:– Industrial composition– Local labour market– Extent and reach of networks– Spatial variations in constraints
Variable Innovation
(Product, Goods)
Innovation (Product, Services)
Innovation (Process, Goods
and Services)
S&E employees/Total 0.256*** 0.244*** 0.054
Other high skills empl/Total 0.077 0.253*** -0.006
Training 0.370*** 0.522*** 0.626***
New Management Technique
-0.097*** 0.093*** 0.191***
New Organisational Structure
0.124*** 0.200*** 0.146***
New Marketing Strategies 0.436*** 0.373*** 0.334***
Collaboration: Local 0.033 0.259*** 0.173***
Collaboration: National 0.396*** 0.271*** 0.393***
Collaboration: Overseas 0.356*** 0.004 0.109*
Obstacles: Finance 0.222*** 0.260*** 0.205***
Obstacles: Knowledge 0.076** 0.098*** 0.071**
Obstacles: Market 0.147*** 0.067*** 0.004
Obstacles: Other -0.085*** -0.008 -0.043
MV Probit Estimates, Fourth Community Innovation Survey, 2004
Variable Innovation
(Product, Goods)
Innovation (Product, Services)
Innovation (Process, Goods
and Services)
R&D expenditure/Turnover 0.004 -0.030 0.012
Regional Public Support 0.083 0.140** 0.079
Government Public Support
0.313*** 0.140* 0.396***
EU Public Support -0.235*** 0.130 -0.099
Claimed R&D Tax Credit 0.497*** -0.117 -0.035
Market: Local reference reference reference
Market: National 0.241*** 0.145*** 0.116***
Market: International 0.588*** 0.141*** 0.256***
Age 0.001 -0.111*** -0.008
Size: Employees (LN values)
0.025*** -0.017* 0.068***
Ownership Structure 0.065** 0.049 0.042
Constant -2.445*** -1.359*** -1.774***
Observations: 13,237
Log Likelihood: -15753.06
MV Probit Estimates, Fourth Community Innovation Survey, 2004, cont.
CIS and CBR
• Analysis of CIS confirms:– Importance of science and engineering employees– No effect of level of R&D expenditure– Importance of training (but CIS does not
disaggregate by type)– Importance of new management techniques (for
services and process innovation)– Importance of national and international
collaborations– Role of tax credits for manufacturing
CIS and CBR
• Additionally analysis of CIS suggests:– A stronger impact of supplying national and
international markets– Regional policy support positive for service
product innovation– Government support positive for manufactured
product innovation and process innovation– EU support negative for manufactured product
innovation
Innovation strategies and absorptive capacity
• Closed Innovation - firms rely on internal resources and the controlled environment of the corporate laboratory.
• Open Innovation - a process where firms use both internal and external sources of ideas and deploy multiple business models to improve corporate performance
• Adoptive Innovation - utilise innovations developed elsewhere
Variable Innovation developed internally
Innovation developed in
collaboration
Innovation adopted
externally
S&E employees/Total 0.488*** -0.222** -0.098
Other high skills empl/Total 0.050 0.163* -0.045
Training 0.473*** 0.419*** 0.319***
New Management Technique
-0.011 0.097** 0.023
New Organisational Structure
0.168*** 0.044 -0.130***
New Marketing Strategies 0.421*** 0.135*** 0.099**
Collaboration: Local 0.101** 0.200*** -0.052
Collaboration: National 0.125** 0.524*** 0.278***
Collaboration: Overseas 0.045 0.071 -0.147*
Obstacles: Finance 0.206*** 0.151*** 0.151***
Obstacles: Knowledge 0.102*** 0.001 0.086**
Obstacles: Market 0.061** 0.040 0.109**
Obstacles: Other -0.066 -0.041 -0.105**
MV Probit Estimates, Fourth Community Innovation Survey, 2004
Variable Innovation developed internally
Innovation developed in
collaboration
Innovation adopted
externally
R&D expenditure/Turnover 0.056* -0.019 -0.603
Local and Regional Public Support
0.122** 0.122** 0.127
Central Government Public Support
0.192*** 0.238*** -0.118
EU Public Support -0.202*** 0.087 -0.037
Claimed R&D Tax Credit 0.392*** -0.154 -0.050
Market: Local Reference Reference Reference
Market: National 0.144*** 0.067 0.099
Market: International 0.429*** 0.161*** 0.047
Age -0.050 -0.044 0.167***
Size: Employees (LN values)
0.036*** 0.032*** -0.015
Ownership Structure 0.052* 0.116*** -0.031
Constant -1.465*** -2.093*** -2.149***
Observations: 13,277
Log Likelihood: -12845.12
MV Probit Estimates, Fourth Community Innovation Survey, 2004, cont.
Open, Closed and Adoptive Innovation
• R&D employees only important for closed innovation• New Management Techniques significant for open
innovators• Local and national collaborations significant for both
open and closed innovators– but the effect is larger for national collaborations– and the effect is larger for open innovators
• Importance of serving international markets
Case Study: Rationale
• To understand the dynamics and interrelationships of absorptive capacity variables
• To provide greater understanding of causal processes
• To explore the role of place, institutions and policies
Case Studies: Innovation Matrix
M1 (100%)M2 (80%)
S1 (70%) M3 (70%) S3 (70%)
M4 (40%)
S2 (60%)
S4 (0%)
S6 (0%) S8 (0%)
S5 (0%) S9 (0%)
S7 (0%) M and S denote manufacturing and services companies respectively % in brackets is R&D employees fraction = (R&D employees/Total Employees)*100
Hig
h R
&D
em
p r
atio
Ze
ro R
&D
em
p r
atio
Collaboration No Collaboration
I II
III IV
Case Studies: Characteristics
I II III IV
Turnover (£mln) 2.1; 3.0 & 3.8 4.5; 9.0; 9.7 & 14 3.2; 7.9 & 16.6 2.2; 3.5 & 14.5
Number of employees
40; 43 & 68 60; 120; 124 & 99 20; 100 & 175 23; 130 & 260
Age (years) 6-17 14-23 21-30 16-22
Location EM & SW SE & SW EM, NW & SE EM, NW & SE
Sectors Semiconductors, Organic Chemicals & Software Supply
Semiconductors, Medical Instruments, Engineering and Architectural Activities
Management, Technical and Software Consultancy and Engineering Activities
Business Services, Software Supply and Computer Related activities
Case Studies: Innovation Behaviour
I II III IVmanufactured product 3 2 1service product 3 2manufacturing process 1service process 1 1manufactured product 1service product 1 1 1 3manufacturing process 2 1service process 1 1 2 2
new to the firm and industry
new to the firm only
Innovation reported
Case Studies: Quadrant I and Quadrant II
• Q1– Important role of external sources for innovation
such as market, customers and suppliers – Company ability to understand and exploit such
sources – Innovation is iterative and interactive
• Q2– Emphasise internal sources for innovation – Innovation seen as linear – interactions with
customers only at the final stages of delivery
Case Studies: Quadrant III and Quadrant IV
• Q3 – Important role for external sources and internal
capability– High innovation expenditure on non R&D activities
eg, training and innovation related ICT• Q4
– Passive reactive innovators – respond to external demands (customers, consultants and technical standards)
Case Studies: Quadrant dynamics Collaboration No Collaboration
M1 S3
M2 S2
S1 M3
M4
S4
S6 S8
S5 S9
S7
Hig
h R
&D
em
p ra
tioZero
R&D
em
p ra
tio
Case Studies Across-Quadrant Results: Absorptive Capacity
Indicators
• Related prior knowledge and staff skills– Founders knowledgeable about the core
technology– Employees have a high general level of education – Use of external and internal training but training
budgets tend to be small – Use of variety of recruitment channels tailored to
the company needs
Case Study Across-Quadrant Results: Absorptive Capacity
Indicators• Organisational structure and management
practices• Job and functional rotation• Quality management practices• Internal knowledge exchange mechanisms• Emphasis on corporate culture• Importance of workplace layout to maximise
learning and knowledge exchange
Case Studies Across-Quadrants Results: Barriers to Innovation
• Finance • Shortage of skilled labour• Location specific factors
– Local labour market– Attractiveness to labour– Infrastructure
• Choice of Location– Importance of residential preference and path
dependence
Case Studies Across-Quadrants Results: Role of Networks
• Local networks– Low rates of participation– Few real benefits other than general awareness
raising• National and international networks
– Provide innovation-related benefits mainly in the form of technical and professional inputs
Case Study Across-Quadrants Results: R&D Tax Credits
• Manufacturing:– General easy access– Benefit is through improving cash flow not
directly influencing the decision to innovate • Services:
– Found it difficult to qualify– Some have resorted to professional
accounting firms to negotiate acceptable settlement
Policy Implications• Different forms of absorptive capacity have different effects on
goods, service and process innovation.• Management practices are important for innovation• Training is important – but the evidence suggest that it is the
training of managers not scientists that is important • R&D tax credit has a positive and significant association with
manufacturing product innovation. • But case studies suggests that the transmission
mechanism is through the impact on cashflow• Networks are important, but their geographical impact varies
across types of innovation.• Policy focus in UK to develop local and regional networks
might be detrimental in some sectoral and regional contexts.
Collaboration and Clusters
• According to Porter (1998) clusters are ‘geographic concentrations of interconnected companies, specialised suppliers, service providers, firms in related industries, and associated institutions (for example universities, standards agencies, and trade associations) in particular fields that compete but also co-operate’
Collaboration and Clusters
• The stress on ‘geographic concentrations’ is misleading• The Porter analysis is based on where firms are, not
what firms do • Local collaborative behaviour is not an important
driver of innovation • National collaborators are consistently more important
than local collaborators• Important to distinguish between ‘bridging and’
bonding’ networks – bridging networks that are outward looking may be more important for knowledge based activity
Further Information
• Centre for Business Research
www.cbr.cam.ac.uk
• Programme on Regional Innovation
www.regionalinnovation.org.uk