Doran and O'Leary (2011) - The Interaction between Internal R&D Performance and External Networking for Innovation: Evidence from the Irish CIS

SLNIW: 2nd International Conference on Networks Learning and Entrepreneurship, Waterford I.T. 7th/8th Dec 2011

The Inter-relationship between R&D Performance and External Networking for Innovation: Evidence from the Irish C.I.S.

Justin Doran & Eoin O’Leary,

Department of Economics

University College Cork

Introduction

This paper estimates the extent to which the interrelationship between R&D spending and external networking contributes to the innovation performance of Irish and foreign-owned businesses.

By providing econometric analysis of the Community Innovation Survey: 2006 to 2008, it contributes to an evaluation of Irish innovation policy.

Begins by profiling policy stance, then positions paper in context of recent literature on Irish innovation, leading to the formulation of a model to be tested.

Goes on to present descriptive statistics and regression results, followed by policy conclusions

2

Irish Policy Context Most important policies that delivered ‘Celtic Tiger’

12.5% corporation tax Targeting high-technology foreign-owned sectors

Objective was ‘Industrialization by Invitation’ with multi-nationals embedding in Irish economy

Policy a success if need for IDA pipeline lessened over time. Little Irish research done on different types of agglomeration

economies (O’Leary, 2007), with exception of Van Egeraat (2006). Reputation and demonstration effects may also be important in

explaining multi-national location (Barry et al , 2003). After 1990s, high cost led to focus on ‘Innovation by Invitation’ - only

‘smart’, ‘knowledge-based’ need apply!! Continued reliance on IDA pipeline – even more important in current

crisis A “very special kind” of micro-state (O’Leary, 2011)

Irish Policy on R&D

Government target to increase R&D by businesses and HEI’s.

Business +5.3% pa

Source: CSO, BERD Releases

Irish Policy on R&D and Networking Incentives for BERD

12.5% corporation tax attractive to all business. R&D is a cost centre – why would foreign businesses locate it

here? 25% R&D tax credit, grants etc eliminate disincentive. Range of additional supports for Indigenous industry

R&D Funds; Competence Centres; Applied Research Enhancement. Business Incentives linked to HERD

Investment in science seen as attractive to industry In addition, tax exemptions for all business on patents/licences

sourced in European Economic Area Range of supports for Indigenous industry working with HEIs

Innovation Vouchers; Innovation Partnerships; Technology Transfer; Spin-outs etc

Repeated emphasis on importance of Networking – Culliton (1992), Enterprise Strategy Group (2004), National Spatial Strategy (2002).

5

Recent Evidence on Irish Innovation Policy Innovation policy criticized for

science-push focus through role of HEI’s focus on high-technology Evidence that external networking with HEIs might have limited

effect (Jordan and O’Leary, 2005, 2008 & 2011 – 184

High-tech) Now is a good time to evaluate innovation policy

Availability of broad-based Community Innovation Survey. Are Some Businesses More Likely to Innovate than Others?

Businesses in 5 sectors (including high-tech) are equally likely Foreign-owned businesses more likely to innovate ‘Dichotomous knowledge sourcing’ - some businesses network

with market agents and others, notable high-tech with non-market agents.

(Doran and O’Leary, 2011 – Irish CIS).

Recent Evidence on Irish Innovation Policy What is Role of External Networking for Innovation and

Productivity? Theory is that interactive learning crucial for innovation

performance External networking affects decision to innovate but not the

exploitation of innovation for productivity gains Important to drill deeper into R&D activity in Irish businesses

(Doran and O’Leary, 2011 – Irish CIS).

What are Private Returns to R&D Spending? For intramural spending Irish-owned businesses significantly

more likely than foreign-owned to introduce new products. Foreign-owned businesses enjoy very high returns to product

innovation for purchase or licence of patents, know-how. Further evidence of dichotomous innovation system suggesting

that ‘one-size-fits all’ policy may not be appropriate.

(Doran, Jordan and O’Leary, 2011 – Irish CIS).

7

R&D, Networking and Innovation

R&D linked to absorptive capacity – increased ability to identify, assimilate and exploit knowledge for innovation (Cohen and Levinthal, 1989).

Research confirms R&D spending has consistently strong positive effects on probability and intensity of innovation (eg Crepon et al, 1998; Roper et al 2008, Doran and O’Leary, 2011)

External networking facilitates access to knowledge from both market and non-market agents.

Varying views in literature regarding the importance of networking relative to internal knowledge generation (Doran and O’Leary 2011; Roper et al 2008).

Standard approach treats R&D and Networking as independent (eg Jordan and O’Leary, 2008).

R&D, Networking and Innovation The main contribution of this paper is to analyse the inter-

relationship between networking and R&D. The following bivariate probit model is estimated

Where IOi is a binary indicator of whether firm i introduced an innovation

RD indicates firm i’s expenditure on R&D Net*RD indicates the relationship between firm i’s expenditure on R&D and

networking.

Predicted values from an estimation of a multivariate probit model for firms’ decision to network are used in equation (1) for Net to control for potential endogeneity.

(1) 1321 hikihkijihijijhih εZη*RDNetφRDφNetφδIO

Are Irish & Foreign-Owned Different?

A Chi2 test is undertaken to assess whether there is a statistical difference in the determinants of innovation for Irish and foreign-owned firms.

This tests the hypothesis that estimating equation (1) for both Irish and foreign-owned firms separately is more efficient than estimating equation (1) for the full sample.

If it is found that separate estimations are more efficient this suggests that grouping Irish and foreign-owned firms together in an econometric model provides incorrect estimates of the model.

This may result in the ‘identification’ of incorrect relationships between variables and the generation of conclusions and policy implications based on flawed results.

The Irish CIS – Innovation measures Product innovation is defined as the introduction of a new or

improved good or service with respect to its capabilities, user friendliness, components or sub-systems.

Process innovation is defined as the implementation of a new or improved production process, distribution method, or support activity for goods or services.

The Irish CIS provides information on whether firms engage forward networks (to customers), backward networks (to suppliers and consultants), horizontal networks (to competitors) and public networks (to universities and public research institutes).

Intramural R&D is defined as creative work undertaken within the enterprise to increase the stock of knowledge for developing new and Improved products and processes.

This paper also considers the relationship between R&D and networking. To account for this effect a multiplicative interaction term is generated which accounts for the simultaneous performance of R&D and engagement in networking.

11

The Irish CIS – Summary Statistics

12

Irish-Owned Foreign-Owned

stat S.D stat S.D

Innovation

Product Innovator (1/0) 26.58 n.a. 45.4 n.a.

Process innovator (1/0) 34.24 n.a. 49.23 n.a.

External Networking

Public interaction (1/0) 4.91 n.a. 7.77 n.a.

Backwards interaction (1/0) 7.77 n.a. 16.41 n.a.

Forward interaction (1/0) 6.09 n.a. 11.78 n.a.

Horizontal interaction (1/0) 2.61 n.a. 4.44 n.a.

Firm Specific Characteristics

R&D (€ per worker) 1402 3042 2381 3821

Employment (no. workers) 63 216 172 307

Source: Irish Community Innovation Survey 2006-08

Effect of inter-relationship between R&D Spending and Networking on the Probability

of Innovation?

R&D

Inter-relationship between R&D and Networking

Public Backward Forward Horizontal

Irish-OwnedProduct +0.26 +0.78 -0.88

Not Significant

Not Significant

Process +0.19 +1.16 -0.65 -0.94 +0.84

Foreign-Owned

Product +0.20

Not Significant

Not Significant

Not Significant

Not Significant

Process +0.09

Not Significant

Not Significant

Not Significant

Not Significant

Implications for Policymakers More evidence of dichotomous innovation systems for Irish and

Foreign-owned businesses. Consistent evidence that R&D matters for all businesses For Irish-owned, strong evidence that interrelationship between

networking and R&D is important Public and horizontal networking act as complements to R&D spending Contrary to Jordan and O’Leary (2008), evidence that HEI networking matters. Backward and forward networking substitutes for R&D spending Cost of R&D forces SME’s to engage in networking (Freel, 2003) Next step to investigate whether agents with whom Irish-owned businesses

network are located within Ireland!! For Foreign-owned no evidence that networking matters as only

R&D spending explains innovation performance Performance of innovation seems to be ‘self-contained’ Other factors such as human capital working with R&D may be important? These firms also spend significant amounts on purchase of licences and patents

– not addressed here!! ‘One-size-fits all’ innovation policy may not be appropriate.

Questions Welcome

[email protected]@ucc.ie

15