SMEs in a Globalised World
SMEs in a Globalised World Survival and Growth Strategies on Europe’s Geographical Periphery
Edited by
Helena Lenihan Department of Economics, Kemmy Business School, University of Limerick, Ireland
Bernadette Andreosso-O’Callaghan Department of Economics and Euro-Asia Centre, Kemmy Business School, University of Limerick, Ireland
Mark Hart Economics and Strategy Group, Aston Business School, Aston University, UK
Edward Elgar Cheltenham, UK • Northampton, MA, USA
© Helena Lenihan, Bernadette Andreosso-O’Callaghan and Mark Hart 2010 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical or photocopying, recording, or otherwise without the prior permission of the publisher. Published by Edward Elgar Publishing Limited The Lypiatts 15 Lansdown Road Cheltenham Glos GL50 2JA UK Edward Elgar Publishing, Inc. William Pratt House 9 Dewey Court Northampton Massachusetts 01060 USA
A catalogue record for this book is available from the British Library Library of Congress Control Number: 2009940737
ISBN 978 1 84844 009 8
02
Printed and bound by MPG Books Group, UK
Contents List of figures List of tables List of contributors Preface Acknowledgements 1
SMEs in a globalised world: conceptual issues Helena Lenihan, Bernadette Andreosso-O’Callaghan and Mark Hart
PART I:
2
3
4
5
6
1
KEY DRIVERS OF A GLOBALLY COMPETITIVE SME SECTOR IN THE EU
Why do SMEs grow? A rejection of Gibrat’s law for Spanish firms (1994–2002) Mercedes Teruel-Carrizosa Access of small firms to knowledge networks as a determinant of local economic development Miren Larrea, Alazne Mujika and Mari Jose Aranguren Innovation behaviour of Spanish fashion manufacturing SMEs José L. Calvo and Angel L. Culebras de Mesa Family-based firms: evidence from the Portuguese furniture and events organisation industries Vitor Braga and Bernadette Andreosso-O’Callaghan Forms of industrial development in Chinese specialized towns and types of challenges to European manufacturing SMEs: an Italian perspective Marco Bellandi and Annalisa Caloffi
PART II:
7
vii viii x xii xiii
19
47
67
87
113
MNEs, SMEs AND INDUSTRIAL DEVELOPMENT
MNE subsidiaries, productivity spillovers and SMEs Rita Buckley v
135
vi
8
9
10
SMEs in a globalised world
Entrepreneurship and inward foreign direct investment in Portugal Natália Barbosa and Vasco Eiriz The large leader firm: good or bad? A case study of a leader firm–supplier relationship Helen McGrath and David Jacobson Dynamics of the SME sector in Ireland: a driver of growth in the Irish economy since 1994? Helena Lenihan, Briga Hynes and Mark Hart
Index
161
181
202
225
Figures 3.1
3.2 3.3 5.1 5.2 6.1
8.1
8.2 8.3
Factor analysis of attitude towards decisionmaking; importance given to management autonomy and geographical distribution of the sales Factor analysis of information systems quality and geographical distribution of the sales Factor analysis of innovation process and geographical distribution of the sales Theoretical model of factors that influence successful succession Funding sources Four groups of specialized towns in Guangdong (2003–05) (means of the seven indices in the four groups of specialized towns) Relationship between net local entry rate and MNEs’ presence for the Portuguese manufacturing and services industry Relationship between net local entry rate and MNEs’ presence for four groups of manufacturing industries Relationship between net local entry rate and MNEs’ presence for different sizes of enterprises
vii
59 61 62 92 101
120
173 175 176
Tables 2.1 2.2
Employees and number of firms (2003) Entry, exit, turbulence and net entry rates for 1999 and 2005 2.3 Percentages of firms in the cohort created in 1994 and observed in 2001 2.4 Average value by type of manufacturing industry 2.5 Average value by type of service industry 2.6 Variables used in the model 2.7 Determinants of firm growth for the whole database (GLS random effects model) 2.8 Impact of territorial variables (region of reference: Madrid) 2.9 Determinants of firm growth for the whole database (GLS random effects model) 2.10 Regional variables (region of reference: Madrid) 3.1 Distribution of the population and sample 3.2 Groups of firms derived from the factor analysis 4.1 Evolution of Spanish fashion manufacturing industry: employment, industrial production index and commercial balance of payments 4.2. Innovating firms in the period 2003–05 4.3 Characteristics of Spanish fashion innovative SMEs 4.4 Share of new or improved products in business turnover 4.5 Main characteristics of Spanish fashion innovation patterns 5.1 Advantages and limits of a family-based firm 5A.1 Characterisation of the firms interviewed 6.1 Indices and source of data 6.2 Distance among the centres of the cluster 7.1 Foreign presence and productivity levels in the Irish indigenous software industry 7.2 Foreign presence and labour productivity growth in the Irish indigenous software industry 7.3 Labour productivity growth for low, medium and high absorptive capacity firms 8.1 Descriptive statistics of variables for the Portuguese manufacturing industry viii
23 24 25 28 30 32 34 36 38 41 55 63
72 76 78 79 81 103 112 118 120 145 147 151 168
Tables
8.2 8.3 9.1 9.2 10.1
Descriptive statistics of variables for the Portuguese services industry GMM estimation of annual local entry rate in the period 1986–2000 Collaboration between leader firm and supplier: Superquinn and Oceanpath Dependency between leader firm and supplier: Superquinn and Oceanpath Estimates from Irish evaluation studies which have derived or incorporated estimates of deadweight as applied to the case of SMEs
ix
168 172 190 191
214
Contributors Bernadette Andreosso-O’Callaghan, Department of Economics and EuroAsia Centre, Kemmy Business School, University of Limerick, Ireland. Mari Jose Aranguren, Orkestra, Basque Institute of Competitiveness and University of Deusto, Spain. Natália Barbosa, School of Economics and Management, University of Minho, Portugal. Marco Bellandi, Faculty of Economics, University of Florence, Italy. Vitor Braga, Middlesex University Business School, UK and Escola Superior de Tecnologia e Gestão de Felgueiras – Instituto Politécnico do Porto, Portugal. Rita Buckley, Department of Economics, Kemmy Business School, University of Limerick, Ireland. Annalisa Caloffi, Department of Economics, University of Florence, Italy. José L. Calvo, UNED, Madrid, Spain. Angel L. Culebras de Mesa, University of Alcala, Madrid, Spain. Vasco Eiriz, School of Economics and Management, University of Minho, Portugal. Mark Hart, Economics and Strategy Group, Aston Business School, Aston University, UK. Briga Hynes, Department of Management and Marketing, Kemmy Business School, University of Limerick, Ireland. David Jacobson, DCU Business School, Dublin City University, Ireland. Miren Larrea, Orkestra, Basque Institute of Competitiveness and University of Deusto, Spain. Helena Lenihan, Department of Economics, Kemmy Business School, University of Limerick, Ireland.
x
Contributors
xi
Helen McGrath, School of Real Estate and Construction Economics, Dublin Institute of Technology, Ireland. Alazne Mujika, University of Deusto, Spain. Mercedes Teruel-Carrizosa, Industry and Territory Research Group, Department of Economics, Rovira i Virgili University, Spain.
Preface This volume brings together a collection of original pieces from leading European industrial economists on the theme of how SMEs, located in what were once considered as being some of the weakest peripheral economies of the ‘old’ EU, have responded to the dual challenges of globalisation and industrial restructuring. The work is based on papers presented at the 9th European Network on Industrial Policy (EUNIP) International Conference which was hosted by the University of Limerick, Ireland in June 2006. Following the conference, we approached a number of participants with an outline of our thinking about this topic and asked them to contribute to this book. The contributions in this volume draw on empirical evidence of some of the traditionally weak peripheral economies of the ‘old’ EU, namely Ireland, Italy, Portugal and Spain, and provide evidence of the processes at work whereby these economies, and particularly the SME sector therein, have been transformed. We develop two interrelated themes in this work; the first examines the main drivers of a globally competitive SME sector in the EU through selected national studies; the second, investigates the relationship between MNEs, SMEs and industrial development (that is, an investigation of the dependent nature of SME growth). The changing nature of the EU economies from a Fordist to a post-Fordist mode of production organisation, characterised by small firms, monopolistic competition (and network firms), as well as flexible specialisation, has created a very different market for SMEs to operate within. All of this implies major challenges and opportunities for SMEs and thus the current volume provides a very timely contribution to a much under-researched area. An important dimension of this volume is that it provides empirical and theoretical contributions on SMEs in both manufacturing and service sectors. We see this work not only as an output but as a mechanism which we hope will encourage further debate by its readers and research endeavours by scholars in this exciting field of research. Helena Lenihan Bernadette Andreosso-O’Callaghan Mark Hart
xii
Acknowledgements The editors would like to acknowledge the valuable contributions of all participants at the 9th EUNIP International Conference, University of Limerick, Ireland, in June 2006. In particular, we would like to thank those who responded to our invitation to participate in this project which sought to bring together collaborations on one of the important themes of that conference – SMEs in a globalised economy. The logistics of putting together this volume have been greatly assisted with the professional and patient support of Liz Blackford (Aston Business School) and Christine Gordon (Economic Research Institute of Northern Ireland) to whom we are highly indebted. Helena Lenihan would also like to acknowledge her hosts at the Centre for Business Research and Wolfson College, University of Cambridge, UK and CSME, Warwick Business School, University of Warwick, UK where she spent time working on this volume during her sabbatical period in 2008.
xiii
1.
SMEs in a globalised world: conceptual issues Helena Lenihan, Bernadette Andreosso-O’Callaghan and Mark Hart
THE GLOBAL ECONOMIC CRISIS – NEW RULES FOR SMEs? Since the end of 2007 the economies of Europe have undergone a number of seismic financial shifts which have transmitted themselves into the business and consumer sectors enforcing recessionary processes which brought to an end the long period of benign growth over the last 15 years. The global economic crisis spinning out of the financial sector in the last two years has abruptly brought into question the nature of the emerging consensus on the processes of globalisation and how they can be harnessed to provide development stimuli in many underdeveloped regions and nations of European economies. This book was originally conceived in the Autumn of 2006 around a series of presentations at the 9th European Network on Industrial Policy (EUNIP) International Conference at the University of Limerick in Ireland which provided an opportunity to document and understand in some depth the ways in which some of the more peripheral economies of the European Union (EU) had managed to transform their economies through the engagement of an indigenous small- and medium-sized enterprise (SME) sector in a global market place. We recognise that the deepening global economic crisis has once again changed the terms of engagement for these economies and their business population and that many commentators and policymakers are struggling to develop policies which remain relevant in a rapidly changing economic context. What we are sure of is that the role of foreign direct investment (FDI) in the current crisis is both simultaneously predictable (retrenchment and cost reduction in the face of falling demand) and unpredictable (financial instability and uncertainty within these organisations leading to speedy 1
2
SMEs in a globalised world
and colossal withdrawal) which creates even greater levels of instability in both the donor and host regions and countries. Further, the risk of a return to national protectionist policies may seriously challenge the competitive position of those SMEs who had managed to overcome the not insignificant barrier of engaging in international marketplaces. However, what we must not lose sight of in the current crisis is the fact that over the last 20 years SMEs in the EU15 have constantly adapted to the challenges of global trends as witnessed by the rise of the BRIC economies (Brazil, Russia, India and China). The world has always been a turbulent place for those businesses seeking to become ‘global’ in their spheres of operation. Consequently, there are many lessons to be learnt from their individual survival and growth strategies and this volume is designed to further deepen our knowledge of these processes as we bring together a series of original research contributions. These include an examination of the drivers of SME growth in a variety of national contexts and the issue of how SMEs cope in an increasingly globalised world and especially the competitive pressures from China. Further, we have sought to investigate the notion of ‘dependent growth’ or how the interaction of SMEs with other firms (for example, multinational enterprises (MNEs) or large firms) has aided performance and competitiveness. Before detailing the nature of the individual contributions we set out some overarching conceptual issues that govern our understanding of the ways in which internationalisation and globalisation have impacted upon the SME sector. We start by discussing the notion of globalisation in some detail and assess the nature of the opportunities and challenges for SMEs in the European economic system and particularly in the EU15. We then proceed with a discussion of the globalised market place for SMEs and describe the way it has changed in the last 10–15 years. SMEs are viewed by many commentators as important engines of economic growth and development in that they inter alia promote job creation, private ownership, act as a catalyst for innovation and develop entrepreneurial skills (Schmögnerová, 2004; van Stel et al., 2007; OECD, 2008; European Commission, 2009). Earlier work by Mytelka (1999), for example, argued that SMEs create the majority of jobs, encourage diversification of economic activities and also make a significant contribution to exports and trade. It is doubtful if the current global economic crisis will affect this assessment unduly although it might constrain the scale at which firms engage in these strategies. Further, from a policy perspective innovation remains largely the work of smaller firms in developed economies via the process of creative destruction which they argue plays a key role in the overall economic growth process. Acs et al. (1997), for example, have claimed that ‘smaller firms are better at creating radical
Conceptual issues
3
innovations because they better protect the innovator’s property rights’ (p. 17). More recently, Breitzman and Hicks (2008) have also argued that small firms are a significant source of innovation and patent activity (that is, as measured by patents per employee) compared with larger businesses. In a similar vein Kirchhoff et al. (2007) argue that entrepreneurial firms are often as innovative, if indeed not more innovative, than their larger firm counterparts. SMEs can survive and grow in a turbulent global economic environment if they are flexible, innovative, customer focused, and both proactive and reactive in their business strategies. Strategically chosen and wellmanaged corporate alliances by SMEs will continue to make it possible for many businesses to overcome some of the disadvantages of being small. As outlined by La Croix (2006) if small firms are to ensure survival and profitability, possible strategies include: ● ● ● ●
specialisation in one aspect of distribution or production; focusing on particular market niches where economies of scale are not a determinant of competitiveness; focusing on flexible service industry specialisation; focusing efforts on economies of speed (facilitated to a large extent by low cost information technology).
How well these strategies can be implemented in a liquidity crisis remains to be seen but as strategic options to management they remain central to coping with the current economic downturn and recession. The increased dominance of service industries will also in all likelihood mean that there will be a dominance of knowledge-based service industries thus making the quality of human capital a key driving factor in terms of ensuring the growth and survival of SMEs in an increasingly globalised world. Further, the presence of large labour abundant economies such as India, China and Brazil (in addition to some of the new EU Member States), where labour costs are significantly cheaper than those in the rest of the EU, mean that in all probability labour intensive activities will increasingly continue to locate to those regions. One of the most optimal ways that SMEs can respond is by focusing on the production of knowledge-intensive goods and services. In many instances, this will involve further enhancement and investment in R&D capacity. Accessing finance in developed industrial economies to undertake R&D investment is the key challenge at the end of the first decade of the twenty-first century which seemed implausible for businesses in these economies in the period before the financial sector imploded.
4
SMEs in a globalised world
DEFINING GLOBALISATION The issue of how to define ‘globalisation’ is one of the most emotive and widely debated topics in economics. As is the case with the concept of ‘sustainable development’ for example, many use the term but few can agree as to its meaning. Here, we argue that globalisation is a multidimensional process. As argued in a report to the Commonwealth Secretariat/ World Bank Joint Task Force on Small States (2000) ‘The process [of globalisation] is driven by continuing innovation and technological change and is associated with the elimination or reduction of national barriers to the global movement of goods, services, capital, technology and certain kinds of labor’ (p. 44). Put simply, we acknowledge that globalisation encompasses a web of connections and relationships between economies, societies and organisations that together comprise the economic system. Globalisation as a concept transcends the discipline of economics, as it is also a political and socio-cultural concept, given that it comprises a process of networking and interconnections involving the integration of people, firms and governments on a global scale. As a process, therefore, it impacts on culture, the environment, on political systems, on individual well-being in societies and on broader economic development and wealth issues. In this volume, however, we confine our discussion of globalisation to that of exploring the economic dimensions and how they impact on the individual firm and its operating environment. From our perspective, globalisation as an economic process is driven primarily by international trade (exports and imports); investment (inward and outward); capital inflows and inter-country labour mobility and overall, is strongly influenced by information technology. As evidence of globalisation in action, we see the following as features of the economic landscape: free flows of information; the breaking down and in some cases removal of technological barriers; prevalence of economies of scale; mergers and acquisitions and the availability and movement of workers globally. We now look in some detail at the opportunities and threats/challenges that are likely to face SMEs as they attempt to survive and grow in an increasingly globalised world. Opportunities for SMEs Globalisation presents many opportunities for SMEs. One of the key opportunities is that of the disappearance of trade barriers. The opening up of national frontiers provides firms, which previously operated on a local/regional/national level, with new foreign market opportunities, thus
Conceptual issues
5
allowing them to survive and grow as this provides an important competitive stimulus. Other dominant opportunities identified in the literature for SMEs emanating from the forces of globalisation include: access to new and/or niche markets; possibilities for the absorption of excess production capacity or output; exposure to international best-practice (for example, with respect to technology); possibilities to exploit economies of scale, scope and volume; means of diversifying risks; the minimisation and sharing of costs (for example, R&D and finance costs) and the optimality of market segmentation (Davenport and Bibby, 1999; Mundim et al., 2000; Narula, 2004; OECD, 2004a, 2004b; OECD, 2008). Threats /challenges for SMEs Although free trade may be seen as something that favours SME activity, it could also be reasonably argued that, given that the elimination of trade barriers, it may also result in increased competition, ‘increasing competition in international markets may have a negative impact on the survival rates of (small) business’ (Verheul et al., 2002: 24). Other threats and challenges which are likely to face SMEs as a result of increased globalisation may include: the high costs that SMEs face in terms of establishing and maintaining foreign distribution and marketing networks and in a related vein the difficulties involved in coordinating complex business to business relationships and supply chains remotely; inadequate protection of property rights; increased competition from imports; entry of new foreign investors into domestic markets and loss of traditional markets to lowerpriced competition from overseas (Acs et al., 1997; Mundim et al., 2000; OECD 2004a; 2004b; OECD 2008). In summary, therefore, given that globalisation implies that firms will compete in a worldwide market (as highlighted by Levitt, 1983; and Porter, 1986), this will present both opportunities and challenges for SMEs. In the chapters that follow, the various contributors explore in much more detail these challenges and opportunities, and indeed coping and development strategies that SMEs in some of the more peripheral countries of the EU have adopted, in an increasingly globalised business operating environment.
THE GLOBALISED MARKET PLACE FOR EU SMEs Up until the summer of 2008 two major economic forces have shaped post-war Europe: one is the changing nature of the mode of industrial
6
SMEs in a globalised world
production with the emergence of SMEs as contestable business actors, and the other is regional integration1 culminating, on a larger scale, with globalisation. First, breaking with decades of conglomerates and cartels, the economies of the EU had evolved into a post-Fordist mode of production organisation, characterised by small firms, monopolistic competition (and network firms), as well as flexible specialisation. In particular, the network firm is a business structure in which a product or service becomes the end result of a synergy involving different firms each highly specialised in a specific stage of the production process (Antonelli, 1988; Robinson and Stuart, 2007).2 This has created a very different market context within which SMEs could flourish and operate. Second, the EU, a well acknowledged pioneer in terms of modern regional integration, has increasingly asserted itself as an important global actor by fostering economic links with the rest of the world. In particular, both economic integration within the EU proper, and the EUs external relations, have been carefully designed so as to be compatible with GATT/WTO liberalisation rules and as a consequence have ultimately fostered globalisation.3 The theme of this book provides, therefore, a symbiosis of these two forces, that is, regional integration on the one hand and evolving modes of production on the other hand. Consequently, its focus on SMEs, located in what may be described as peripheral locations in the EU, is embedded in the current wave of globalisation. The different chapters in this volume aim at investigating the two-way relationship existing between globalisation and SMEs. First, by analysing the way in which the survival and growth trajectory of SMEs based in the EU periphery is influenced by globalisation, and second by clarifying the ways in which SMEs have responded to the global competition challenge. It is intended to illustrate the role that SMEs have played in transforming these peripheral economies. With regard to the first relationship, the way in which SMEs respond to a more globalised market (or at lower level, to the challenges of the Single European Market) implies analysing two broad dimensions in the internationalisation (or globalisation) strategy of firms. One dimension refers to the ‘geographical spread’, or the number of countries in which the firm is operating; the other to the ‘geographical depth’, or the operation mode of the firm (Johanson and Vahlne, 1977). Firms venture into foreign markets either through trade or through direct investment. According to a recent survey commissioned by the European Commission, only 8 per cent of all EU SMEs are involved in exports, compared with 19 per cent for large firms, that is, those employing more than 250 people. Only 5 per cent of all EU SMEs have internationalised by way of direct investment abroad, either through greenfield projects or through joint ventures, against 20 per cent for firms with more than 250 employees (Gallup Organization,
Conceptual issues
7
2007).4 The results of the survey show that the export propensity increases with the size of the SME, and that the export propensity is an inverse relationship of the size of the country, with SMEs from larger countries, such as France, exporting less. Trade also involves the purchase of inputs abroad (outsourcing), with 12 per cent of the inputs (excluding labour) of an average EU SME being purchased abroad.5 Although EU SMEs are still marginally present in international markets compared with their larger counterparts, recent trends denote their increasing engagement in cross-border trade. In their study on Swedish SMEs from Kalma County, Hilmersson and Sandberg (2007) show the positive impact of the 5th EU enlargement of May 2004 on the internationalisation, through trade, of Swedish firms, in particular in the Baltic States and in Poland. Other peripheral countries of the EU, such as Portugal and Ireland, display above EU average shares of SMEs involved in foreign direct investment. The ‘stage-model’ explaining the internationalisation of SMEs, whereby the SME develops its capabilities in its home market before venturing abroad, is very much disputed however by the work of Andersson and Wictor (2003) who focus on the internationalisation strategy of ‘born-global’ Swedish firms. Fernandez and Nieto (2006) highlight the importance of ownership on the internationalisation of Spanish SMEs over the period 1991–99, with family firms being the least likely to go global. For the largest part of all EU SMEs, internationalisation actually means overwhelmingly ‘Europeanisation’; some 70 per cent of the exporting EU SMEs indicate another EU country as the main export destination, with Germany ranking first, followed by France, Spain, the Netherlands, Italy and the UK (Gallup Organization, 2007). This can be viewed as epitomising a first step in the internationalisation process of EU SMEs. Deeper economic integration in the EU, with the completion of the Internal Market and the Lisbon Agenda, have all stimulated cross-border deals, not only for large firms but also for their smaller counterparts. However, a puzzling result of the Gallup Organization 2007 survey is the view by EU SMEs that a major business constraint in the EU has been, and still is, the limited purchasing power of customers. This was mentioned as being the major impediment to growth by 46 per cent of the surveyed SMEs. This implies that limited growth prospects in the EU market might play as an incentive for venturing beyond the EUs borders. At present, Asia and America/Oceania are the targets in only 9 per cent and 10 per cent respectively of all business partnerships formed by EU SMEs. Comparatively, some 17 per cent of large EU firms have such business partnerships in Asia. This second step in the internationalisation process of SMEs can only fruitfully take place if, and only if, a number
8
SMEs in a globalised world
of deterrent factors are dealt with. The emerging economies (particularly those of Asia) are notorious for shielding behind a number of tariff and non-tariff barriers. The EU SMEs that were surveyed list the lack of knowledge of foreign markets and the barriers to entry as the two most prominent barriers to exporting. With regard to the second relationship, that is, the way SMEs can shape globalisation, the potential role of SMEs in the transformation of EU economies has recently been a very much debated issue in EU circles. The 2003 Sapir Report emphasised the increasingly crucial role of small firms in an enlarged Single European Market (SEM), and the authors of this report see SMEs as a one of the key drivers of EU growth in the future (Sapir et al., 2003). But, as argued above, although the supply-side conditions are on the whole much more favourable to the EU SMEs in recent years than they ever were before, sluggish and now rapidly falling demand in the SEM represents an important constraint for these firms. This problem can in part be tackled by the rejuvenation of the SME sector through its greater linkage with economies outside the EU border. According to the Gallup Organization survey (2007), the EU SMEs who are involved in foreign markets through FDI, report a net positive impact of their foreign activity on their domestic employment. Labour costs (in the EU) are ranked only as the third reason for venturing into foreign markets through FDI. The major reasons for business partnerships abroad are indeed the proximity to final customers (evoked in 17 per cent of the cases), and the proximity to other firms that purchase the SMEs’ inputs (in 12 per cent of the cases). This is even more so the case for micro firms, an element that tends to demonstrate that micro firms are able to insert themselves in the global systems of production (GSP), at least in relative terms. It follows that EU SMEs can only play a crucial role in revitalising economic growth in the SEM if, and only if, a number of prerequisites are met. EU policy rightly sees the Lisbon Strategy (reviewed and renewed in 2005) as encompassing two complementary and inseparable facets: the new global Europe strategy (the external agenda) and the SEM (the internal agenda). However, little attention is given to the SMEs, as crucial actors, in the ‘market access strategy’ of the new global Europe strategy. Attempts were made under, for example, the Asia Invest programme, launched in 1998, to stimulate the internationalisation of European and Asian SMEs, through greater integration between the two regions. Given that competitive pressures have intensified in the EU, and that domestic demand fails to provide an adequate stimulus to economic growth, a recentring of the Lisbon Agenda and of its two facets, towards EU SMEs might be appropriate.
Conceptual issues
9
STRUCTURE OF THE BOOK This volume brings together a number of leading European industrial economists to address the theme of the role of SMEs in helping the economies of the EU meet the challenges of globalisation and restructuring. There are nine chapters and they are divided into two parts: Part I – key drivers of a globally competitive SME sector in the EU; and Part II – MNEs, SMEs and industrial development. What is significant about these empirically-based contributions is that almost all concern EU economies that were once considered the weakest of the European economies – namely, Ireland, Spain, Portugal and Italy. Accordingly, they may provide important lessons for the ways in which the SME sector can potentially contribute to the overall development of traditionally weak economies and inform policy debates for SME development in some of the new EU Member States. The chapters on Spain are important as we seek to shed some light on the processes at work which underlie the observation that in 2005 Spain had at 4.2 per cent, after the United Kingdom 6.4 per cent, the second highest proportion of high growth firms of all the countries in EU15 for which data was available.6 Part I of the book contains five chapters which address the ways in which the SME sector has been developing in recent years in three EU economies: Spain, Portugal and Italy. These chapters open the discussion by analysing key issues brought about by globalisation and by proffering responses to these new challenges. The growing importance of SMEs in the Spanish economy is explored through an investigation of three-related issues. First, the determinants of growth in the manufacturing and service sectors (Chapter 2 by Teruel-Carrizosa); second, the importance of the connectivity of small firms in Spain, and in particular the Basque country, to knowledge networks as a determinant of local economic development (Chapter 3 by Larrea et al.); third, the role of innovative behaviour in the future survival of SMEs in the Spanish fashion manufacturing industry (Chapter 4 by Calvo and Culebras de Mesa). Chapter 2 proposes both a test of Gibrat’s Law and an investigation into industry-specific differences on firm growth. The study concludes in favour of the rejection of Gibrat’s Law in favour of small firms. There is a convergence process of firm size, with small firms growing faster than their larger counterparts, and the author finds that this convergence process is more rapid in the manufacturing than in the services sector. Chapter 3 looks at the specific case of SMEs in the Basque Country (Spain) and suggests that networking activities are necessary to allow these SMEs to respond to the challenges of globalisation. This chapter presents a conceptual model on
10
SMEs in a globalised world
how networks can boost SMEs’ competitiveness, through innovation and knowledge management. The authors find that firms with between 10 and 50 employees in the Basque Country seem to deal with globalisation in isolation. Chapter 4 analyses the challenge brought about by globalisation, and in particular by the ‘made in China’ syndrome, on the European fashion industry, a particular concern for Spain, where jobs destruction affects specific regions and SMEs. As reiterated by the authors, innovation is the essential response to these challenges, for it helps to increase competitiveness and maintain employment. However, the authors warn against a fragile innovation strategy among Spanish fashion SMEs, with for example only 25 per cent of all Spanish fashion manufacturing firms (mostly SMEs) being innovators. Consequently, the authors call for policies to encourage firm’s innovation activities. The final two chapters in Part I provide further evidence on the ways in which the SME sector can contribute to the national competitiveness agendas. The role of the changing nature of family-based firms in two Portuguese industries – furniture manufacture and events organisation – provides an opportunity to investigate the ways in which small firms have to adapt to changing national and international competitive pressures (Chapter 5 by Braga and Andreosso-O’Callaghan). The main focus of Chapter 5 is on cooperation (networking activities) in the early years of the firm’s existence, specifically in the case of the furniture and events organising industries in Portugal. Using a qualitative methodology, the authors find a number of relevant differences in terms of the process of business start-up, and particularly with regard to the funding strategies of the nascent firm. Furniture entrepreneurs, who are mostly rural-based, rely essentially on strong ties (that is, they engage in strong and familybased networking activities), although a generational change is noted by the authors. In contrast, event organisation entrepreneurs are seen as modern Portuguese entrepreneurs. Their firms tend to be scattered over the major cities, and the typical entrepreneur in this new industry is more educated than those of the furniture industry. The authors also show that, when faced with modern business challenges such as rising competition, the firms seek new forms of relationships. Finally, an investigation of the role of industrial development through ‘clusters’ in Guangdong Province in China is presented to provide lessons for the future development of the once renowned Italian industrial districts who are under the competitive pressure from SMEs in a rapidly industrialising Chinese economy (Chapter 6 by Bellandi and Caloffi). Many Italian (and indeed, European) SME-based production systems in industrial districts have been hit by a growing competitive challenge coming from these emerging countries. With this in mind, the authors propose to analyse cluster
Conceptual issues
11
forms of industrial development in China’s Guangdong Province, and to inform on business and policy reactions in Italian industrial districts facing this new international competitive challenge. The authors conclude that Chinese specialised towns can be best understood along the lines of European models of local development such as industrial districts and SMEs clusters, and that these represent new challenges to the established Italian districts. Part II of the volume contains four chapters which explore the relationship between SMEs and multinational enterprises (including both FDI and nationally-owned). The aim here is to present some of the latest research evidence on the important question whether MNEs produce spillover and linkage effects and, if so, what their overall contribution is to both the development of a competitive SME sector and industrial development in the host country. Buckley (Chapter 7) examines the issue of MNE subsidiaries, productivity spillovers and SMEs in the case of Ireland. Ireland’s growth and exporting performance since the mid-1980s is primarily attributed to the presence of foreign-owned MNE subsidiaries especially in high technology sectors. This chapter empirically investigates the impact of MNE subsidiaries on the productivity performance of SMEs using the Irish software industry as a case study. Buckley details that the experience of the indigenous software industry in Ireland supports the view that indigenous SMEs in a peripheral region of the EU can successfully survive and indeed prosper alongside a strong MNE sector due to learning and demonstration effects between foreign and domestic firms. The author does however emphasise that benefits such as those alluded to above are neither universal nor automatic. The chapter provides interesting lessons vis-à-vis the degree to which SMEs benefit from productivity spillovers due to the presence of large foreign-owned MNE subsidiaries. The Irish case is also explored in two of the other chapters in Part II as follows: McGrath and Jacobson (Chapter 9) present a case study of a leader firm–supplier relationship in the fish processing sector in North Dublin which highlights the ways in which seemingly ‘peripheral’ small firms can develop an international competitive position with strategic alliances. The chapter details how the development of a relationship between these two firms leads to much success and growth for the SME. The authors proceed to argue that the leader firm and its supplier enjoy a collaborative, reciprocal and bilateral arrangement. McGrath and Jacobson do caution however, that not all leader firm–supplier relationships are necessarily beneficial to the small firm and its development. More generally, Lenihan et al. (Chapter 10) explore the dynamics of the SME sector in Ireland with a particular focus on the extent to which
12
SMEs in a globalised world
enterprise development and entrepreneurship can be associated with the rapid growth of the Irish economy in the 10 years from 1994 (the widely accepted date of the start of Ireland’s phenomenal economic growth rates). The profiling of SMEs undertaken in the chapter leads the authors to conclude that the SME sector has grown considerably since 1994. Consistent with international evidence, they also report that small firms create and destroy jobs at a higher percentage rate compared with their larger firm counterparts. However, one of the most interesting findings, based on previously unpublished VAT data on firm births and deaths over the period 1988 to 2005, is that there does not appear to be any evidence to support the conclusion that Ireland’s rapid economic growth was not driven by a rise in new venture creation. Further, the analysis also shows that the rapid growth period in the Irish economy can also not be seen as having provided a stimulus to increased gross or net business birth rates. Their overriding conclusion is that economic growth can happen at a national level without a concomitant increase in business start-up activity. This is a particularly interesting finding vis-à-vis small firm policy a component of which often seeks to stimulate business start-up activity with the overall objective of stimulating economic growth through the process of new business creations. Lenihan at al. emphasise that the collection and on-going maintenance of reliable datasets is a necessity if rigorous SME research and subsequent policy improvements are to be undertaken. SMEs are seen as having the potential to play an important contributing role in terms of the future growth dynamics of the Irish economy, particularly in the face of globalisation and its associated opportunities and challenges. The authors also emphasise that for SMEs in Ireland, increased engagement in exporting strategies and that development of SMEs in high-value added and services sectors is crucial. In view of the current economic downturn, which is impacting on Ireland more severely within the EU27, the development of small firm activity will have an important role to play in any Irish economic recovery strategy but not as the sole driver of economic growth and wealth creation. As other chapters in this volume have illustrated, much of the growth in the SME sector has been dependent in the past on their connectivity to supply chains managed and controlled by the FDI sector. Finally, the interrelationship between entrepreneurship and inward investment is examined by Barbosa and Eiriz (Chapter 8) who analyse empirically whether MNEs operating in Portugal have a positive impact on local entrepreneurial activity. Outside high-technology sectors and large firms the analysis suggests a negative impact of inward FDI on the net creation of local firms. The authors interpret the result as suggesting a positive impact of MNEs on industrial restructuring by forcing the exit of less efficient firms.
Conceptual issues
13
NOTES 1. Regional integration needs to be understood as referring to the dismantling of (primarily economic) barriers between countries belonging to geographically delineated broad areas such as Europe, North America and Asia. 2. According to the early work by Antonelli (1988) the major advantages of these networks are that they allow firms to be more flexible and innovative, because of the short production runs. Robinson and Stuart (2007) look at strategic alliances in the biotechnology industry and use the notion of ‘alliance network’. 3. A recent case in point is the ‘new Global Europe Strategy’ (CEC, 2006; and CEC, 2007), which implies stronger links, through ‘new generation’ bilateral FTAs, between the EU and a number of selected countries, and which, in the eyes of the European Commission, is purely complementarity with the multilateral WTO system. 4. The survey was conducted in 2006/2007 over the EU25, plus Iceland, Norway and Turkey, and it involved 16 339 SMEs. 5. Interestingly, the results of the study show a negative and statistically significant correlation between size of an economy and proportion of inputs purchased abroad; but no statistically significant impact of firm size on backward linkages. 6. The definition of high growth firms follows that of the OECD (www.oecd.org/std/ industry-services) and is explained in detail in EUROSTAT-OECD (2007) Business Demography Manual. The employment-based definition in simple terms is there needs to be at least 20 per cent average annual growth in employment over three years with the condition of at least 10 employees in the base year. Therefore, the proportion reported here for Spain and the UK refers to the 2002–05 period.
REFERENCES Acs, Z.J., R. Morck, J.M. Shaver and B. Yeung (1997), ‘The internationalization of small and medium-sized enterprises: a policy perspective’, Small Business Economics, 9 (1), 7–20. Andersson, S. and I. Wictor (2003), ‘Innovative internationalisation in new firms: born-globals – the Swedish case’, Journal of International Entrepreneurship, 1 (3), 249–76. Antonelli, C. (1988), ‘The emergence of the network firm’, in C. Antonelli (ed.), New Information Technology and Industrial Change: The Italian Case, Amsterdam: Kluwer Academic Publishers, pp. 13–32. Breitzman, A. and D. Hicks (2008), ‘An analysis of small business patents by industry and firm size’, report for the Small Business Administration (SBA), Office of Advocacy, Haddonfield, NJ: SBA. Commission of the European Communities (CEC) (2006), ‘Global Europe: Competing in the World – A Contribution to the EU’s Growth and Job Strategy’, COM /2006/ 567, 4 October, Commission of the European Communities, Brussels. Commission of the European Communities (CEC) (2007), ‘Global Europe: a stronger partnership to deliver market access for European exporters’, COM (07) 183 Final, 18 April, Commission of the European Communities, Brussels. Commonwealth Secretariat (2000), ‘Small states: meeting challenges in the global economy’, report prepared for the Commonwealth Secretariat/World Bank Joint Task Force on Small States, Washington DC, April.
14
SMEs in a globalised world
Davenport, S. and D. Bibby (1999), ‘Rethinking a national innovation system: the small country as “SME”’, Technology Analysis and Strategic Management, 11 (3), 431–62. European Commission (2009), ‘Think small first – considering SME interests in policy-making including the application of an “SME test”’, report to the European Commission Enterprise and Industry Directorate-General, Brussels. EUROSTAT-OECD (2007), Manual On Business Demography Statistics, Brussels and Paris: European Commission. Gallup Organization (2007), ‘Observatory of European SMEs’, technical and evaluation report, Flash Eurobarometer Series no. 196, Budapest: European Commission. Fernandez, Z. and M.J. Nieto (2006), ‘Impact of ownership on the international involvement of SMEs’, Journal of International Business Studies, 37 (3), 340–51. Hilmersson, M. and S. Sandberg (2007), ‘EU enlargement effects on international trade in the Baltic Sea region – the case of exporting/importing SMEs from southern Sweden’, paper to the 9th Vaasa Conference of International Business, Vaasa, Finland, 19–21 August. Johanson, J. and J. Vahlne (1977), ‘The internationalisation process of the firm – a model of knowledge development and increasing foreign market commitments’, Journal of International Business Studies, 8 (1), 23–32. Kirchhoff, B.A., S.L. Newbert, I. Hasan and C. Armington (2007), ‘The influence of university R&D expenditures on new business formations and employment growth’, Entrepreneurship: Theory and Practice, 31 (4), 543–59. La Croix, S.J. (2006), ‘Globalization and SMEs: a comment on three Asian experiences’, University of Hawaii, Manoa Department of Economics working paper 200603. Levitt, T. (1983), ‘The globalization of markets’, Harvard Business Review, 61 (3), 92–102. Mytelka, L.K. (1999), ‘Globalization and investment: a learning and innovation approach’, presentation to the DRUID Conference on National Innovation systems, Industrial Dynamics and Innovation Policy, Rebild, Denmark: 9–12 June. Mundim, A.P.F., R. Alessandro and A. Stocchetti (2000), ‘SMEs in global market: challenges, opportunities and threats’, Brazilian Electronic Journal of Economics, 3 (1), accessed 17 April 2009 at venus.unive.it/stocket/SMEs%20 in%20global%20market.pdf. Narula, R. (2004), ‘R&D collaboration by SMEs: new opportunities and limitations in the face of globalisation’, Technovation, 24 (2), 153–61. Organisation for Economic Co-operation and Development (OECD) (2004a), ‘Promoting SMEs for Development’, presentation at the second OECD conference of ministers responsible for small and medium-sized enterprises, Promoting Entrepreneurship and Innovative SMEs in a Global Economy: Towards a More Responsible and Inclusive Globalisation, Istanbul, Turkey: 3–5 June. OECD (2004b), ‘Facilitating SMEs access to international markets’, presentation at the second OECD conference of ministers responsible for small and mediumsized enterprises (SMEs), Promoting Entrepreneurship and Innovative SMEs in a Global Economy: Towards a More Responsible and Inclusive Globalisation, Istanbul, Turkey: 3–5 June. OECD (2008), ‘Removing barriers to SME access to international markets’, Industry, Services and Trade Series, 2008 (5), 1–217.
Conceptual issues
15
Porter, M.E. (1986), ‘Changing patterns of international competition’, California Management Review, 28 (2), 9–40. Robinson, D.T. and T.E. Stuart (2007), ’Network effects in the governance of strategic networks’, Journal of Law, Economics and Organization, 23 (1), 242–73. Sapir, A., P. Aghion, G. Bertola, M. Hellwig, J. Pisani-Ferry, D. Rosati, J. Viňals and H. Wallace (2003), An Agenda for a Growing Europe – Making the EU Economic System Deliver, Brussels: European Commission. Schmögnerová, B. (2004), Statement of Executive Secretary at the 6th Forum on Best Practice in Development of Entrepreneurship and SMEs in Countries in Transition: The Romanian and Slovak Experiences, United Nations Economic Commission for Europe, Geneva, 31 March, accessed 17 April 2009 at www. unece.org/press/execsec/2004/bs040331.htm. van Stel, A., D.J. Storey and A.R. Thurik (2007), ‘The effect of business regulations on nascent and young business entrepreneurship’, Small Business Economics, 28 (2–3), 171–86. Verheul, I., S. Wennekers, D. Audretsch and R. Thurik (2002), ‘An eclectic theory of entrepreneurship: policies, institutions and culture’, in D. Audretsch, R. Thurik, I. Verheul and S. Wennekers (eds), Entrepreneurship: Determinants and Policy in a European–US Comparison, Boston, MA: Kluwer Academic, pp. 11–81.
PART I
Key drivers of a globally competitive SME sector in the EU
2.
Why do SMEs grow? A rejection of Gibrat’s law for Spanish firms (1994–2002) Mercedes Teruel-Carrizosa
INTRODUCTION One of the main concerns of policymakers and researchers has been the analysis of the engines of economic growth. Researchers agree that the dynamism and the capacity of small firms is a key variable that influences economic development (Carre and Thurik, 1998; Robbins et al., 2000). There is an extensive international literature related to small firms and their importance in terms of job creation (Acs and Audretsch, 1989; Carre and Thurik, 1998; Fariñas and Moreno, 2000; Calvo, 2006). Firm growth has been one of the most studied topics in the economic literature. Whether firm performance responds to the firms internal and external changes or random factors, the economic literature has tried to search the leitmotiv of firm growth. The reason is that firm growth affects the market structure where firms operate. Whether small firms grow faster is not just a mere causality. There are a large number of contributions noting that firms are highly flexible with possibilities to grow and capacity to increase their size (Nooteboom, 1994). However, there are few studies showing the differences between manufacturing and service industries as well as the reasons for those different patterns. From all these dimensions, the firm growth literature studies the relationship between firm size and growth. Several reasons justify the importance of such analysis. First, it is a strategy for firm survival in the long run. Second, the knowledge of firms’ growth patterns will allow policymakers to apply the appropriate policies to increase employment, which has been a crucial issue in Spain from 1998 to 2008. Moreover, the firm growth process will determine the distribution of firms and market concentration. Given the current context of internationalisation and globalisation, the firm growth process has been an acutely felt process in terms of its survival. Our purpose is to analyse Gibrat’s Law for the Spanish firms 19
20
SMEs in a globalised world
between 1994 and 2002. This study makes several contributions to the literature related to firm growth: first, we contribute to the scarce literature about Spanish firms’ growth; second, the panel data covers the medium term; third, we distinguish between firms in the manufacturing and service industries. Gibrat’s Law suggests that the expected increase of a firm size in each period is proportional to the current size of the firm. This forms the basis of our analysis. The competitive environment and the market structural changes oblige us to revise the assumptions of Gibrat’s Law. These changes give small firms the chance to find their own path to grow. As a consequence, a priori we should expect that it is small firms that will have more opportunities to increase their size in order to survive. The majority of the literature analyses the behaviour of manufacturing firms, while there are only scarce contributions to service industries (Audretsch et al., 2004). There are two main reasons why the differences between these two industries are of interest. First, each industry has specific characteristics, which may lead to different firm sizes (Acs and Audretsch, 1989). Second, although the literature indicates that the initial characteristics of each firm can influence long-run firm behaviour, authors such as Görg et al. (2000) have found that industry characteristics have a different affect on the behaviour of small and large firms. Hence, we aim to analyse the impact of these different characteristics on firm growth, with particular attention paid to differences between the two industries. Our results show that Gibrat’s Law is rejected in favour of small firms. This means that small firms tend to grow faster than larger firms in both industries; however, there are differences in firm growth between the manufacturing and service industries. These features appear when analysing the differences between the internal and external characteristics affecting the firm’s growth. This suggests that policymakers trying to enhance the relationship between the manufacturing and service industries should consider the specific characteristics of those industries. With the aims mentioned above, the layout of this study has been structured as follows: the next section summarises the literature related to the growth of firms; afterwards, we review the methodological contributions in the literature; the next section describes the data of our database; this is followed by the presentation of the econometric methodology and the empirical estimation; the last section presents the conclusions.
LITERATURE This section presents a review of the literature related to firms’ growth. First, we briefly analyse the evidence related to Gibrat’s Law and the
Why do SMEs grow?
21
models related to the Theory of Learning. Finally, we highlight a gap in the literature, which is represented by the analysis of the service industries. Empirical Evidence of Gibrat’s Law and the Theory of Learning Stochastic growth assumes that firm growth is a random process. Gibrat’s Law has been a point of interest in order to determine the observed skewed firm size distribution. Gibrat (1931) postulated that firm growth is independent of firm size. That means that all firms have the same probability to grow.1 Consequently, in the market there will be a high number of small firms coexisting with a small number of large firms.2 Sutton (1997), Caves (1998) and Geroski (1999) have pointed out the stylised regularities of firm demography. In particular they find that: (a) the rate of growth declines with the initial size of the firm; (b) entry and exit are closely related with the growth-size relationship, small firms dominate entry and exit declines with size; and (c) growth and exit rates decline with age. Early studies from Hart (1962) and Prais (1976) in the case of the UK, as well as Ijiri and Simon (1977) for the USA accepted Gibrat’s Law. The reason is that Gibrat’s Law provides a good explanation to the skewed distribution of firm sizes observed in the market. However, Mansfield (1962) pointed out some interesting results. He stated that Gibrat’s Law is relevant for surviving firms with size exceeding the minimum efficient scale (MES),3 while the smallest surviving firms tend to have a positive relationship between firm growth and size. Empirical evidence gathered since the 1990s has produced results inconsistent with Gibrat’s Law. While some studies accept the presence of Gibrat’s Law, others reject its hypothesis of independence between firm growth and its initial size in favour of a more rapid growth of small firms. The post-entry performance of firms has also attracted attention in the theoretical literature. Jovanovic (1982) presents a model where firms enter, exit and grow. Jovanovic characterises a model with passive learning, which means that firms cannot modify their unknown initial efficiency level. Finally, Gibrat’s Law is only supported for those new firms that exceed the MES, while smaller firms present a negative relationship between size and growth. Later, Ericson and Pakes (1995) and Pakes and Ericson (1998) developed a model with firm dynamics and firm-specific uncertainty. In this case, firms are heterogeneous and their survival depends on investment, on the behaviour of other firms, on industry characteristics and on external shocks. Consequently, there is an interaction between all the agents of the model. The result will be active learning whereby firms can modify their initial efficiency level through investment.
22
SMEs in a globalised world
There are three recent Spanish studies that need to be pointed out. Fariñas and Moreno (2000) analyse the growth of 2188 Spanish manufacturing firms between 1990 and 1995 using non-parametric methods.4 First of all, they estimate the firm’s growth for all firms according to their size and age. Their results show that size has a negative impact on growth among the surviving firms. Consequently, Gibrat’s Law is rejected. Controlling the mean-reversion bias, their results show that meanreversion does not explain the negative relationship between firm growth and size in the case of surviving firms. Moreover, they show that growth rates of smaller surviving firms diminish abruptly. Their main contributions are that it is one of the few studies using Spanish data, and that they use a non-parametric approach which relates a firm’s age and size with the variance in size and the probabilities of failure. Correa et al. (2003) estimate firm growth in the Canary Islands with data from the Commercial Performance Information Bureau5 between 1990–96. They study firm growth using different measures (assets, added value and incomes). Their results show a positive impact of insularity6 on firm growth and reject Gibrat’s Law as small firms grow faster than large ones. Finally, Calvo (2006) analyses whether small, young and innovating firms display faster growth levels than other Spanish firms over the period 1990–2000. Controlling for sample selection with a Heckman’s equation, his results show that old firms grow less than young ones, and that innovation has a strong positive impact on the firm’s survival and on its employment growth. Furthermore, he finds a significant and positive relationship between firms operating in high technological sectors and firm growth. Finally, innovation appears as a crucial determinant for firm survival. The conclusions from this literature are that there are factors other than the initial size of the firm that determine the growth of a firm. Consequently, we include other explanatory variables such as the location, the external activity and the legal status of the firm in our empirical specifications, which capture other factors that explain the firm growth process. Firm Growth and the Service Industries Although there has been prolific literature focused on the firms in the manufacturing sector, recent contributions studying the service sector have been scarce. Audretsch et al. (2004) analyse the Dutch hospitality sector showing that the different subsectors behave heterogeneously. Oliveira and Fortunato (2006) reject Gibrat’s Law for Portuguese firms in the manufacturing and service sectors; their results show that firms in
Why do SMEs grow?
Table 2.1
23
Employees and number of firms (2003) Manufacturing
Employees Number of firms Employees per firm
Services
Number
%
Number
%
2 532 269 152 915 16.56
37 13
4 391 072 992 338
63 87 4.42
Source: Derived from data drawn from Encuesta Industrial and Encuesta Anual de Servicios (Spanish Statistics Institute, Madrid) (1994–2002).
the service sectors behave in the same way as those in the manufacturing sectors and the authors reject Gibrat’s Law. Audretsch et al. (1999a) argue that firms in the service sectors will perform differently in comparison with those in the manufacturing sectors. In fact, service industries grow more slowly than new manufacturing firms. This is because in the service sectors, scale economies are smaller than those found in the manufacturing sectors, and firms are not constrained to attain a given MES for survival. An inferior level of scale economies in the service sectors implies that firms will have less of an incentive to increase their size. Therefore few empirical studies compare the performance of firms in the manufacturing sectors with those in the service sectors. Because of this gap in the literature, our objective is to compare the performance of firms in the two industries during the same period of time and from the same source of information.7 The analysis of different industries is justified because those industries present some crucial differences. For instance, manufacturing industries generally have a higher MES and higher investment levels. Downsized firms in a manufacturing industry therefore find it difficult to survive and grow. New firms in the service industries may remain under a lower threshold because the scale economies in the service industries are lower than in the manufacturing industries.8 In order to isolate those differences, Table 2.1 shows the relationship between the number of employees and number of firms in the Spanish manufacturing and service industries in 2003. These results show that over 60 per cent of employees work in the service industries. Table 2.1 also presents information about the number of firms in each sector. As we can see, 87 per cent of firms belong to the service industries. From Table 2.1 we can observe that the percentage of workers is substantially higher than the percentage of firms in service industries. Therefore, the MES of service firms is smaller than that of firms in the manufacturing sector.
24
Table 2.2
SMEs in a globalised world
Entry, exit, turbulence and net entry rates for 1999 and 2005 (per cent) 1999 Entry rate
Exit rate
2005 Turbulence rate
Entry rate
Exit rate
Turbulence rate
Total 0 workers 1–5 workers 6–9 workers 10–19 workers More than 20
9.52 14.26 9.52 6.03 4.44 2.27
9.35 17.39 7.60 4.07 2.89 1.53
Manufacturing 18.87 7.96 31.65 16.38 17.11 5.86 10.10 3.58 7.32 2.62 3.80 1.32
7.48 13.18 6.76 3.98 2.86 1.42
15.44 29.56 12.63 7.56 5.48 2.74
Total 0 workers 1–5 workers 6–9 workers 10–19 workers More than 20
15.24 18.07 12.38 7.71 6.05 4.64
13.98 18.07 9.34 4.48 3.44 2.72
Services 29.22 13.75 36.14 19.48 21.71 8.14 12.19 5.52 9.49 3.92 7.36 2.32
9.50 12.23 7.20 4.21 3.09 1.70
23.25 31.71 15.34 9.72 7.00 4.03
Source: Derived from data drawn from Directorio Central de Empresas (Spanish Statistics Institute, Madrid) (1994–2002).
In fact, much of the literature is related to the small size of new firms (Audretsch, 2002; Segarra et al., 2002; Arauzo and Segarra, 2005). It is a stylised fact that entrepreneurs start downsized businesses because: (a) they cannot obtain financial support;9 (b) young firms have high failure levels and, as expectations are unknown, entrepreneurs may decide not to invest a large amount of money; and (c) they misunderstand the real investment needed to take advantage of scale economies. This statement is corroborated by the number of employees per firm (MES): firms in the manufacturing sector have an average of 17 employees and those in the service sector have an average of four. The analysis of the dynamics of both industries also highlights those structural differences. Table 2.2 classifies the rates of entry, exit and turbulence for different sizes of firms and it shows the dynamics for each type of sector between 1999 and 2005 according to firm size groups. In both years, the entry and exit rates decrease as the firm size increases. The relationship between the turbulence rate and firm size is also negative. These results are in agreement with those in the literature on firm demography.10 Moreover, the entry and exit rates are higher for the service industries regardless of
Why do SMEs grow?
Table 2.3
25
Percentages of firms in the cohort created in 1994 and observed in 2001
Manufactures Services
Equal size
Changing size
Failures
20.72 23.81
25.21 17.33
54.07 58.86
Source: Derived from data drawn from Directorio Central de Empresas (Spanish Statistics Institute, Madrid) (1994–2002).
size or year. Turbulence is therefore higher in the service sector than in the manufacturing sector. Having analysed entries and exits, we will now analyse the evolution of a cohort of firms created in 1994. Specifically, we will analyse the period between 1994 and 2001 and examine their growth pattern at the end of this period in order to determine whether firms have remained in the same group size or whether they have joined a different group size. Table 2.3 shows that more than half of the firms created in 1994 had disappeared by 2001.11 This means that firms suffer from a high infant mortality rate, which is in agreement with the literature (Audretsch and Mahmood, 1995; Audretsch et al., 1999b; Segarra and Callejón, 2002; Segarra et al., 2002). However, there are several differences between the manufacturing and the service sectors. First, firms in the service sector are more likely to disappear from the market: 58.86 per cent of firms in this sector had failed by the end of the evaluation period. This may be related to the lower sunk costs and more short-term strategies for facing unemployment situations in the service sector. Second, the MES of service firms is usually lower than in manufacturing firms so service firms have a higher turnover rate. To sum up, there are clearly considerable differences between these two types of industries. Not only are there more firms in the tertiary sector but there are also more workers. It is therefore justified to differentiate between the two sectors. Our interest, however, lies in determining the effect of these differences on the growth trends of the firms.
ESTIMATING GIBRAT’S MODELS In this section we summarise the different estimation procedures used in the literature to estimate Gibrat’s Law. Gibrat (1931) presents a model where the current size of a firm depended on the firm size in the previous period (t21) plus a random error. From this initial point, contributions
26
SMEs in a globalised world
have evolved towards an estimation of firm growth, measured as the difference in size between two periods in logarithms, as a function of the firm’s size in the initial period.12 This can be shown as follows: DlogSi,t 5 ai 1 bi logSi,t21 1 ei,t
(2.1)
where Si,t is the size of firm i in the period of time t, and ei,t is an error term. The majority of the studies suppose that the intercept and the slope are identical for all the firms (that is, ai = aj, bi = bj). Equation (2.1) would be: DlogSi,t 5 a 1 blogSi,t21 1 ei,t
(2.2)
In the case that b < 0, firms converge in the long run to a size equal to a/b. In equilibrium, there are differences from the long-run size, but they are transitory. The process of transition towards the equilibrium proceeds at a speed of b. Equation 2.2 shows that there will be convergence if b < 0, hence there will be a process of mean reversion. In the case where b > 0, there will be a divergent process where firms continue growing, hence larger firms will be relatively larger from the smaller counterparts. Finally, if b = 0, Gibrat’s Law is accepted, since firm growth would be independent of the firm size. Thus small and large firms have the same probability to grow. Consequently, there would not be a process of convergence because firm growth would depend on random shocks. Lazarova et al. (2003) analyse the growth of 147 large UK firms which survived between 1955 and 1985. The authors point out that Equation 2.2 assumes convergence to the same steady state (the convergence would be towards the average of the group). Similarly, the b parameter would be an estimate for the whole group of firms. An alternative would be to apply a fixed effects panel to correct for unobservable and invariable determinants of the size in equilibrium.13 Another problem is the assumption that any change in firm size is a transition towards equilibrium, and that the growth of firms which have reached the equilibrium size (a/b) is due to random shocks. Lazarova et al. (2003) propose Equation 2.3 which is the difference between the size of firm i and j: Ddijt ; D (logSi,t 2 logSj,t) ; bdijt21 1 (ei,t 2 ej,t)
(2.3)
When there are no differences in the intercept (ai = aj), Equation 2.3 shows whether variations in size are due to differences in size in t−1. Firms i and j will converge in the long run if dijt is equal to zero when t approaches to
Why do SMEs grow?
27
infinity. The advantage of this approach is that it does not model trends in growth and differences in sizes can be random or deterministic. The authors find evidence of weak reversion and emphasise two main conclusions: (a) the difference between firms is permanent; and (b) the speed of convergence in traditional models is sensitive to how heterogeneities are handled, that means that heterogeneities in the steady state size or in the speed of convergence can generate biased estimations. They indicate that Equation 2.2 is useful for active firms which do not deviate excessively from the equilibrium, and suggest applying Equations 2.2 and 2.3 to analyse the convergence in the long run.
DATA DESCRIPTION In this section we present the database, Sistema de Análisis de Balances Ibéricos (SABI). We also describe our sample selection procedure and provide some descriptive statistics. Database The SABI database offers information from Spanish and Portuguese firms that present their accounts in the Business Register. The database contains information on more than 550 000 Spanish firms and 50 000 Portuguese firms since 1994. It covers more than 95 per cent of Spanish firms that present their accounts in the business register. This database has a number of advantages. It covers a large number of firms and contains detailed information on a wide range of firm characteristics. It contains information on the same firms over a relatively long period of time (1994–2002)14 and it allows us to study firm growth over the medium term. Sample Selection Our sample contains 139 922 Spanish firms, of which 68 281 are in the manufacturing sector and 71 641 firms are in the service sector, over the period 1994 to 2002. All firms have a common feature which is that they survived until 2002. However, some firms entered the market before the first period of observation (1994), while others have entered the market since. To overcome problems in estimation caused by missing data, we only include firms in our sample that survived the time period examined that had non-missing information. In this way we have constructed an unbalanced panel data without missing data.15 Also, to take full advantage of
28
SMEs in a globalised world
Table 2.4
Average value by type of manufacturing industry No. of firms observ.
Food and beverages Textile and leather products Wood and cork Paper products and media Chemical manufactures Rubber and plastic products Other nonmetallic products Metal products Machinery and equipment Electric and optic apparatus Transport equipment Furniture Note: Source:
No. of Employees
Income Value Sales per Age of per added per employee firm employee employee
40 280
33.66
145.32
24.43
142.98
12.80
41 484
22.10
88.08
17.97
87.43
9.01
16 220 37 617
17.64 21.23
72.12 79.57
19.20 24.35
71.65 78.49
9.01 9.79
11 493
58.34
135.69
33.99
133.59
16.01
14 350
35.19
92.05
26.86
91.38
11.64
19 775
35.50
90.64
26.49
90.02
11.67
62 648 18 822
22.92 28.76
70.94 86.65
22.88 26.30
70.50 85.76
9.84 11.80
12 522
48.77
99.18
25.82
96.77
11.19
7 194
124.24
106.24
27.42
104.97
12.83
22 858
17.45
62.83
18.45
62.42
8.99
The values of income, value added and sales refer to thousands of euros. Derived from the SABI database (1994–2002).
the panel nature of the data, we include firms with observations in at least two consecutive years. Data Description One of the objectives of this study is to investigate any differences in firm growth between firms in manufacturing and those in the service sector. As a first step in doing this, we summarise these key variables separately within each sector (Tables 2.4 and 2.5).
Why do SMEs grow?
29
In manufacturing (Table 2.4), about 20.5 per cent of observations relate to the metal products industry. Moreover, the metal products industry has one of the smallest average firm sizes (with 23 employees) together with textile and leather products (with 22 employees), wood and cork (with 18 employees) and furniture (with 17 employees). The relationship between the number of active firms in an industry and the firm size necessary to be efficient is well-known in the literature (Segarra et al., 2002). If a firm needs a high number of employees to enter the market in a particular industry, then the barriers to entry are high, so there will be few firms. In our case, the majority of firms concentrate in the aforementioned industries and their average firm size is relatively small, which indicates that those industries have low barriers to entry. Firms in food products and beverages and chemical manufactures have on average a higher income per employee than the other industries. A pattern can be identified between sales per employee and income per employee within all manufacturing industries. The value of both indicators is practically identical for all industries. Firms in chemical manufactures have the highest average value added per employee, at 34 000 euros, while textile and leather products firms, wood and cork and furniture have the lowest at 18 000 euros, 19 000 euros and 18 000 euros respectively. The final column in Table 2.4 shows that the chemical manufacturing industry has the oldest firms with an average firm age of 16 years. Furniture, as well as the wood and cork industries have the youngest firms with an average firm age of nine years. Table 2.5 focuses on the service sectors, and it shows that the majority of firms are in motor vehicles (28.3 per cent), hotels and restaurants (22.2 per cent) and transport (26.9 per cent). Relatively few firms are in the research and development industry (0.5 per cent). In terms of the number of employees, the largest firms on average are in telecommunications (122 employees) and the smallest ones are in motor vehicles and wholesale trade (13 and 11 employees respectively). If we compare income per employee, firms in motor vehicles and wholesale trade have the highest average value (163 000 and 254 000 euros respectively), while firms in the hotels and restaurants sectors have the lowest results in the service category (49 000 euros per employee). Those results are common, and are to be expected, since these are typically the industries with lower value added per employee given their relatively high labour-intensity. As expected, firms in the finance and renting industries have the highest values in terms of value added per employee (85 000 euros and 70 000 euros respectively), while those in telecommunications on average add a value of 5000 euros per employee. Finally, the youngest firms are found in telecommunications and computer activities
30
SMEs in a globalised world
Table 2.5
Average value by type of service industry No. of firms
Motor vehicles Wholesale trade Hotels and restaurants Transport Telecommunications Finance Renting Computer activities R&D Note: Source:
Sales Age of Value per firm added employee per employee
No. of employees
Income per employee
79 167 17 713 61 967
13.43 10.85 27.63
163.43 253.68 49.14
20.76 27.05 19.53
160.61 251.54 47.17
10.75 9.00 8.34
75 176 6 049
26.99 121.61
137.34 93.62
28.21 4.59
135.19 90.34
9.85 5.74
8 895 10 659 18 349
37.24 12.71 30.36
115.29 136.34 89.45
85.42 69.61 22.26
118.64 132.26 87.45
8.58 7.98 5.90
1 316
49.75
89.86
33.48
79.56
8.16
The values of income, value added and sales refer to thousands of euros. Derived from the SABI database (1994–2002).
with an average firm age of six years. Again, this is to be expected as these industries have seen their competitiveness increase during the last decade. In general, the comparison of the firms in the manufacturing and service sectors indicates that those in the service sectors are smaller and younger. However, they have a better performance in terms of income per employee, value added per employee and sales per employee. The final conclusion is that manufacturing and service sector firms are different, which may suggest that they enjoy different rates of growth. Our aim in the remainder of this study is to analyse differences in the post-entry performance of firms in each sector.
EMPIRICAL ESTIMATION Given the above sections, our starting point is that the pattern of firm growth depends on the industrial sector. That means that there are differences in the growth process of a firm in the service sector in comparison to the one in the manufacturing sector. This section presents the estimation and the main empirical results.
Why do SMEs grow?
31
Estimation The prevalent equation used to estimate Gibrat’s Law in the literature is Equation 2.2. However, this estimation supposes that firms behave identically, which in our case assumes that firm growth is identical for firms in the manufacturing and service sectors. However, Audretsch et al. (2004) show that firms from the manufacturing and service sectors behave differently. For this reason, we introduce two lagged size variables: the first one identifies the lagged size of firms in the manufacturing sector (taking a value of 0 for firms in services) and the other captures the impact of the lagged size of firms in services (taking a value equal to 0 for firms in the manufacturing sector). Therefore our main equation takes the following form: DlogSi,t 5 ai 1 bi,MANUF logSiMANUF,t21 1 bi,SERV logSiSERV,t21 1 ei,t (2.4) Furthermore Equation 2.4 will include four groups of variables as shown in Table 2.6. The first group encompasses characteristics such as the lagged firm size and age. The coefficients on the lagged size variables for firms (SiMANUF,t−1 and SiSERV,t−1) will determine whether or not Gibrat’s Law is accepted. In particular, a value equal to 0 will accept the Gibrat’s Law, which means that firm growth does not depend on the initial size of the firm. The age and its square show the effect of experience on firm growth. While Age represents the way the firm evolves when time lapses, Age2 expresses how the maturity process affects firm growth. Those variables are closely related to the Theory of Learning (Jovanovic, 1982). In this sense, Jovanovic argues that the youngest surviving firms will have a higher growth rate. The second vector measures the external activity of the firm. Dummy variables identify firms that export products or services, import products or services, and those that do both (Exp, Imp and ExpImp), and they examine their impact on firms’ growth trends. The third set of variables measures the internal organisation of the firm and its legal status. The variable Holding measures whether the firm has consolidated and unconsolidated accounts relative to those firms with unconsolidated accounts. In addition, limited liability firms are compared with joint stock companies (JStockC) and with firms with another legal status (Others). Finally, the fourth vector captures the effect of firm location. In particular, dummies identify firm location in one of the 19 Spanish autonomous regions (AR1 . . . AR19) to capture the impact of locating in a particular region. We compare firms in other regions to those in the Autonomous Region of Madrid.
32
SMEs in a globalised world
Table 2.6
Variables used in the modela
Age Age2
Individual characteristics Logarithm of the size from the previous period only for manufacturing (with a value of 0 for services) Logarithm of the size from the previous period only for services (with a value of 0 for manufacturing) Logarithm of the age Logarithm of the age squared
Exp Imp ExpImp
External activity Dummy identifying if the firm exports Dummy identifying if the firm imports Dummy identifying if the firm both exports and imports
SiMANUF,t−1 SiSERV,t−1
Internal organisation and legal status Dummy identifying firms with consolidated and unconsolidated accounts Dummy identifying joint stock company Dummy identifying if the firm is not a joint stock company or a limited liability company
Holding JStockC Others
AR1. . .AR19
Note:
a.
Location variables Dummies identifying firm location in one of the Spanish autonomous regions
See Table 2.8 for a description of location variables.
We introduce each set of variables separately to estimate their effect on firm growth. The final model to be estimated has the following specification: DlogSi,t 5 ai 1 b1iMANUF logSiMANUF,t21 1 b2iSERV logSiSERV,t21 1 b3ilogAgei,t 1 b4ilogAge2i,t 1 b5ilogExpi 1 b6i logImpi 1 b7i logExpImpi 1 b8i Holdingi
(2.5)
1 b9i JStockCi 1 b10iOthersi 1 b11i ARi 1 ei,t We apply random effects estimates. Although the Hausman test does not accept the hypothesis that the coefficients in the fixed and random effect models are similar for the different specifications, the random effect models are more efficient since they incorporate information across
Why do SMEs grow?
33
individual firms as well as across periods. The panel data estimation has been used frequently in the literature. Recently, authors such as Chen and Lu (2003) and Resende (2004) have used panel data regression to estimate the post-entry firms’ performance. Results The results of the econometric estimations are presented in Table 2.7. Each column presents the results from introducing a further set of covariates, with specification 6 being the most complete. The results show that Gibrat’s Law is rejected since the coefficients on the lagged size variables are statistically significant for firms in both manufacturing and services (SiMANUF,t−1 and SiSERV,t−1). Moreover, the coefficients are negative which implies that smaller firms grow more rapidly than the larger ones over the observed period. These results are in concordance with recent evidence from the UK (Goddard et al., 2006), Portugal (Oliveira and Fortunato, 2006), Germany (Audretsch and Lehmann, 2005) and Spain (Calvo, 2006). These results are consistent with Jovanovic’s (1982) model since there is bounded efficiency. That means that there is a range of sizes where firms can produce at efficient levels. Second, age does not have a clear and consistent pattern, although being significant in nearly all cases, the sign varies across specifications. The negative sign relating to the variable Age suggests that a decrease in the firm growth indicates diminishing returns to learning. The positive sign of the variable Age2 is due to the fact that those firms surviving for such a long period of time may have some strategies or characteristics which enhance their probability to survive and grow. If one looks at the external activity of the firm, those firms that engage in exports or imports have significantly higher growth rates than those with neither. Pfaffermayr (2004) and Voulgaris et al. (2003) confirm that there is a positive relationship between exports and firm growth process. Furthermore, we find an additional growth premium associated with having both imports and exports – the coefficient on the ExpImp variable is also positive and significant. Consequently, those firms which participate in other markets buying or selling products can reach higher growth rates. We suppose that those firms must have a larger level of efficiency to be able to compete in larger international markets. Those firms belonging to a Holding experience a significantly higher level of growth. This is likely to be the case because they belong to multinational firms or holdings with more capacity to grow and diversify their investments. This diversification gives the opportunity to the firm of increasing its capacity of growth without having to take so many risks as
34
Holding
ExpImp
Imp
Exp
Age2
Age
SiSERV,t−1
SiMANUF,t−1
Table 2.7
−0.1624 (0.0009)* −0.1511 (0.0008)*
Specification 1 −0.1800 (0.0010)* −0.1709 (0.0009)* 0.0053 (0.0001)*
Specification 2 −0.1826 (0.0010)* −0.1874 (0.0010)* −0.0004 (0.0002)*** 0.0001 (0.00001)* 0.0698 (0.0045)* 0.0841 (0.0050)* 0.2306 (0.0038)*
Specification 3 −0.1895 (0.0010)* −0.1937 (0.0010)* 0.0003 (0.0002) 0.0001 (0.00001)* 0.0694 (0.0045)* 0.0873 (0.0038)* 0.2236 (0.0038)* 0.7273 (0.0168)*
Specification 4 −0.2045 (0.0010)* −0.2045 (0.0010)* −0.0034 (0.0002)* 0.0001 (0.00001)* 0.04310 (0.0044)* 0.0565 (0.0050)* 0.1841 (0.0038)* 0.7187 (0.0168)*
Specification 5
Determinants of firm growth for the whole database (GLS random effects model)
−0.2061 (0.0010)* −0.2061 (0.0010)* −0.0034 (0.0002)* 0.0001 (0.00001)* 0.0428 (0.0045)* 0.0548 (0.0050)* 0.1841 (0.0038)* 0.7222 (0.0168)*
Specification 6a
35
0.3823 (0.0019)* 0.4251 414 123 163938.98 (0.0000)
0.3995 (0.0022)* 0.4367 185926.65 (0.0000)
0.3683 (0.0019)* 0.4438 179651.08 (0.0000)
Notes: a This equation includes location variables (see Table 2.8). Standard deviation in parentheses. * significant at the 1% level, ** significant at 5% and *** significant at 10%.
R2 N Hausman test
Constant
Others
JStockC
186199.43 (0.0000)
0.4072 (0.0022)* 0.4401
185529.74 (0.0000)
0.1793 (0.0026)* −0.0127 (0.0102) 0.4237 (0.0022)* 0.4397
186197.70 (0.0000)
0.1816 (0.0027)* −0.0068 (0.0102) 0.4226 (0.0033)* 0.4404
36
Table 2.8
SMEs in a globalised world
Impact of territorial variables (region of reference: Madrid) Whole database Coefficient
AR1–Andalusia AR2–Aragon AR3–Asturias AR4–Balearic Islands AR5–Canary Islands AR6–Cantabria AR7–Castile-La Mancha AR8–Castile-Leon AR9–Catalonia AR10–Ceuta AR11–Valencia AR12–Extremadura AR13–Galicia AR15–Melilla AR16–Murcia AR17–Navarra AR18–Basque Country AR19–Rioja R2 N
Standard deviation
−0.0066 −0.0077 −0.0064 0.0183 0.0913 0.0662 −0.0167 −0.0181 −0.0053 −0.0244 0.0107 −0.0263 0.0314 −0.0931 0.0298 0.0397 0.0165 −0.0053
(0.0040)*** (0.0053) (0.0070) (0.0067)* (0.0081)* (0.0134)* (0.0055)* (0.0052)* (0.0034) (0.0587) (0.0039)* (0.0101)* (0.0055)* (0.0531)*** (0.0069)* (0.0081)* (0.0051)* (0.0105) 0.4251 414 123
Notes: Standard deviation in parentheses. * significant at 1%, ** significant at 5% and *** significant at 10%.
a small firm. In this sense, Delmar and Davidsson (1998) do not find any clear pattern between the group affiliation and firm growth.16 Finally, we find that joint stock companies (JStockC) enjoy significantly higher levels of growth in comparison with limited liability companies. The coefficient on the variable Others (which represents general partnership, limited partnership, cooperatives and associations) has a negative impact although this is not significant. Due to the importance of the location variables, Table 2.8 shows the coefficients obtained in the estimation of specification 6 from Table 2.7. The aim is to examine the different impacts of the geographic location of firms on the firm growth process. A central location in a major commercial capital may be crucial in accessing labour skills, information sources and client bases. Technological advantages which can be easily imitated and rapidly diffused within services are less important than within manufac-
Why do SMEs grow?
37
turing. Hence we might expect location to have a different impact on firm growth. We should remember that the region of reference is Madrid, so all our comparisons will be relative to this Spanish region. In this sense, Madrid has a concentration of the majority of the governmental services and has been an important city in terms of influences. In which regions do firms enjoy the highest growth? In the pooled specification regions with positive coefficients we find Balearic Islands, Canary Islands, Cantabria, Valencia, Galicia, Murcia, Navarra as well as the Basque Country. The largest impact is found in the Canary Islands. As Correa et al. (2003) show, the insularity of the Canary Islands can be highly significant in achieving higher growth. In contrast, firms located in Melilla have the lowest growth, with an estimated coefficient of −0.0931. To sum up, the location has different impacts on firm growth depending on the industry. This difference may be due to the fact that regions specialise in economic activities. While they can exert a considerable positive influence in the manufacturing sector, they have a negative impact in services. The Effect of Industry Characteristics However, the previous specification forces the coefficients and the other covariates to be equal for firms in the two sectors, which may not be realistic – there is no reason why we should expect the impact of exporting goods and products to have the same impact on growth for a firm in manufacturing as for a firm in the services sector. To overcome this, we estimate the same equations distinguishing between characteristics in terms of the manufacturing and services sectors. The reason is that the joint estimation of the manufacturing and services sectors can introduce a bias in the estimations (Table 2.9). The new variables are interactions of firm and external characteristics with dummy variables identifying whether the firm belongs to the manufacturing or services sector. Those new variables aim to control for the difference between the effect of those characteristics and firm growth. To begin with, the impact of firm size on firm growth seems to be the same for the manufacturing and service industries for those estimations which introduce the legal status and the regional dummies (specifications 5 and 6). Concretely, the magnitude of the effects increases when introducing new variables. The results highlight that the coefficient of the lagged firm size in manufacturing industries is larger than in service industries. This result is significantly different from the results obtained in Table 2.7, in particular this means that small firms in the manufacturing sector grow faster than those in the services sector. This effect can be closely related to the fact that
38
Specification 2 −0.1873 (0.0010)* −0.1624 (0.0011)* 0.0084 (0.0001)* 0.0003 (0.0001)***
Specification 1
−0.1624 (0.0009)* −0.1511 (0.0008)*
−0.2086 (0.0011)* −0.1612 (0.0012)* 0.0050 (0.0002)*** −0.0073 (0.0003)* 0.00002 (0.00001)* 0.0002 (0.00001)* 0.0579 (0.0050)* 0.1095 (0.0110)* 0.0853 (0.0062)* 0.0781 (0.0085)* 0.2530 (0.0043)*
Specification 3 −0.2320 (0.0012)* −0.1526 (0.0013)* 0.0033 (0.0003) −0.0031 (0.0003)* 0.00004 (0.00001)* 0.0001 (0.00001)* 0.0630 (0.0050)* 0.0962 (0.0110)* 0.0909 (0.0062)* 0.0752 (0.0085)* 0.2590 (0.0044)*
Specification 4 −0.2479 (0.0013)* −0.1632 (0.0013)* −0.0007 (0.0003)** −0.0063 (0.0003)* 0.0001 (0.00001)* 0.0001 (0.00001)* 0.0309 (0.0050)* 0.0711 (0.0110)* 0.0518 (0.0062)* 0.0519 (0.0085)* 0.2085 (0.0044)*
Specification 5
Determinants of firm growth for the whole database (GLS random effects model)
ExpImpMANUF
Imp SERV
Imp MANUF
Exp SERV
Exp MANUF
Age2SERV
Age2MANUF
AgeSERV
AgeMANUF
SiSERV,t−1
SiMANUF,t−1
Table 2.9
−0.2483 (0.0013)* −0.1640 (0.0013)* −0.0007 (0.0003)* −0.0063 (0.0003)* 0.0001 (0.00001)* 0.0001 (0.00001)* 0.0311 (0.0050)* 0.0722 (0.0110)* 0.0516 (0.0062)* 0.0499 (0.0085)* 0.2088 (0.0044)*
Specification 6a
39
0.3823 (0.0019)* 0.4251 414 123 163938.98 (0.0000)
0.4072 (0.0022)* 0.4345 184630.37 (0.0000)
178340.12 (0.0000)
0.1171 (0.0078)*
0.3757 (0.0019)* 0.4406
Notes: a This equation includes location variables (see Table 2.10). Standard deviation in parentheses. * significant at 1%, ** significant at 5% and *** significant at 10%.
R2 N Hausman test
Constant
OthersSERV
OthersMANUF
JStockCSERV
JStockCMANUF
HoldingSERV
HoldingMANUF
ExpImpSERV
183716.78 (0.0000)
0.1825 (0.0059)* 0.4310
0.0998 (0.0078)* 0.2990 (0.0058)* 0.1665 (0.0061)*
182478.88 (0.0000)
0.0710 (0.0078)* 0.2995 (0.0058)* 0.1573 (0.0061)* 0.2185 (0.0038)* 0.1419 (0.0038)* −0.0494 (0.0123)* 0.0480 (0.0185)* 0.2041 (0.0059)* 0.4321 182829.76 (0.0000)
0.0713 (0.0078)* 0.3001 (0.0058)* 0.1580 (0.0061)* 0.2192 (0.0036)* 0.1428 (0.0038)* −0.0493 (0.0123)* 0.0471 (0.0185)* 0.1807 (0.0138)* 0.4325
40
SMEs in a globalised world
the MES is larger for the manufacturing industries than for the services sector. Therefore, manufacturing firms must grow in order to achieve the medium efficient size. This fact suggests that the impact of firm size is not independent from the industry where the firm operates; so within manufacturing industries, the initial endowments of the firm may be a crucial variable for their post-entry behaviour. With reference to the learning effect, the experience in the market (Age) has a greater negative impact on service industries than on manufacturing industries. That is, the older the firm, the lower its growth rate. The results in the literature are somewhat mixed. On the one hand, Fariñas and Moreno (2000) find that there is a negative relationship between firm growth and age; Lotti et al. (2003) find a negative relationship which disappears over time. On the other hand, authors such as Calvo (2006) find a positive relationship between firm size and firm growth. The export–import effect shows a strong linkage between exportation and firm growth in the case of service industries. Conversely, firms importing have a larger impact on firm growth in manufacturing industries than in the service industries. However, the largest differences appear with those firms exporting and importing simultaneously. Although both coefficients are positive, their impact is larger for manufacturing firms. This result is in concordance with the results of Calof (1994) and Machado and Mata (2000) who observed that large firms are more intensively involved in international trade. Furthermore, there is a significantly larger coefficient for manufacturing firms than for service firms in the case of the holding variable. The explanation for those results differ according to the sector: in the manufacturing sector, firms need a large MES to survive in order to be competitive; and in the services sector, firms may have great business opportunities without having to belong to a group of firms. Finally, the legal status variable leads to conflicting signs depending on the sector. On the one hand, manufacturing industries have a large positive effect when they are a joint stock company; on the other hand, there are opposite impacts on firm growth when firms have a different legal status than being a joint stock company and limited liability company (manufacturing firms have a negative effect and service firms have a positive effect). Finally, when analysing the effect of regional dummies on firm growth (see Table 2.10), we find that some of the regions, which previously presented a negative sign, show a positive sign (Castile-La Mancha, CastileLeón, Catalonia, Ceuta, Extremadura, Melilla and Rioja). Our results can be explained by the importance of firm characteristics depending on the sector. So, as we have argued theoretically, location is a key issue in explaining firm growth.
Why do SMEs grow?
Table 2.10
41
Regional variables (region of reference: Madrid) Whole database Coefficient
AR1–Andalusia AR2–Aragón AR3–Asturias AR4–Balearic Islands AR5–Canary Islands AR6–Cantàbria AR7–Castile-La Mancha AR8–Castile-León AR9–Catalonia AR10–Ceuta AR11-Valencia AR12–Extremadura AR13–Galicia AR15–Melilla AR16–Murcia AR17–Navarra AR18–Basque Country AR19–Rioja R2 N
Standard Deviation
−0.0202 −0.0170 −0.0311 0.0264 −0.0112 0.1194 0.0420 0.0230 0.0289 0.0219 0.0180 0.0044 0.0302 0.0205 0.0269 0.0504 0.0176 −0.0036
(0.0126) (0.0014) (0.0186)*** (0.0179) (0.0159) (0.0168)* (0.0133)* (0.0126)*** (0.0034)** (0.0135) (0.0134) (0.0152) (0.0132)** (0.0598) (0.0140)* (0.0146)** (0.0152)* (0.0153) 0.4325 414 123
Notes: Standard deviation in parentheses. * significant at 1%, ** significant at 5% and *** significant at 10%.
The results presented in Table 2.9 suggest that small firms achieve higher growth rates, even when controlling for characteristics belonging to different sectors. They also indicate that manufacturing firms grow faster than service firms. A possible explanation is the difference between the MES of manufacturing and service industries. Furthermore, there appears important differences in the regional explanatory variables when taking into account the duality between sector and firm characteristics.
CONCLUSIONS The internationalisation and globalisation context has changed the competition among industries in the market and especially in Spain where productivity growth has been stagnating between 1998 and 2008. The
42
SMEs in a globalised world
objective of this study was to apply the stochastic firm growth theory to Spanish firms. Employment growth generation and the dynamics of SMEs are key issues for the economic situation of a country like Spain, where unemployment and the lack of incentives to firm growth have been crucial. Our results reject Gibrat’s Law and we find that there is a negative relationship between growth and firm size: the smallest firms grow faster than the largest firms. The main results can be summarised as follows. First, Gibrat’s Law is rejected in favour of small firms. Small firms grow faster and, consequently, there will be a trend of convergence of sizes in the industry. Our results show that those SMEs which are able to survive will be able to grow at a higher rate than large firms. Second, variables such as belonging to a group of firms, being a Joint Stock Company or having external activity exhibit a positive effect on firm growth. Third, there is heterogeneous behaviour between sectors when considering the sectorspecific characteristics. Fourth, the locational effects play an important role in explaining differences between firms in the manufacturing and the services sectors. In particular, location exerts an effect on the evolution of firm growth, which may explain why firms in some regions obtain better performance. The results contribute to the international and Spanish literature by focusing on firms in both the manufacturing sector and in the service sector. Our purpose was to investigate the presence of different industryspecific differences on firm growth. In this sense, results show that there is a significant difference in growth between manufacturing and service firms. Policymakers have the objective of increasing economic activity and employment in particular regions. Subsidies to firms can be a tool to achieve this objective. However, policymakers must consider the efficiency of their programmes. This study gives some clues on which characteristics indicate potential firm growth. Those indicators are dynamic and they should be reviewed periodically in order to assess the impact of incentives and barriers to firm growth in the market. One way in which the study would find an interesting extension is by using data on micro firms, something that was beyond the limits of the present database. Our results are useful for policymakers as it gives some information about the determinants of firm growth. First, although small firms show higher potential to grow and they are agents which introduce dynamicity in the market, we should also bear in mind that Spanish firms are on average smaller than European ones, so Spanish policymakers should promote their growth. Second, locational externalities may influence the firm growth behaviour. In that sense, policymakers should also provide those facilities necessary to create a conducive environment.
Why do SMEs grow?
43
Third, it is desirable that surviving small firms are able to increase their size; however the internationalisation and globalisation context has increased the likelihood that small firms have to find a market niche and this has increased competitive pressures. As a consequence, policymakers will also have to consider the international competitive environment and the industrial policies in order to apply the most appropriate policies.
NOTES 1. 2. 3. 4. 5. 6. 7. 8.
9.
10.
11.
12. 13. 14. 15. 16.
Due to the inconsistency between empirical evidence and the neoclassical theory of firm growth convergence Geroski (1999) argues in favour of Gibrat’s Law. This phenomenon would explain the skewed distribution of firms in an industry. The minimum efficient scale is the point where economies of scale are fully exploited. Their source of information is the Encuesta sobre Estrategias Empresariales. Commercial Performance Information Bureau. The database includes all firms belonging to non-financial sectors. The peculiar characteristics of the Canary Islands means that firm growth is higher in the Canary Islands than in the rest of Spain. We consider that the homogeneity of the data is crucial to examine Gibrat’s Law. Different sources of information may reduce comparability and therefore may substantially distort the results. There is an important type of behaviour that we do not take into account. There is evidence that some entrepreneurs create their own business as a strategy against temporary unemployment. Therefore, as soon as they find an interesting job, they may not continue the business. Storey (1994) pointed out that entrepreneurs find it difficult to create businesses because of liquidity constraints. Moreover, the lending policy of banks seems to be unrelated to the personal characteristics of entrepreneurs, which have a significant impact on firm growth. For international results, Geroski (1995) and Caves (1998) reviewed the stylised facts and results on firm demography. For Spain, see Callejón and Segarra (1999) and Segarra et al. (2002). However, all of these contributions analysed manufacturing industries only. We consider that a firm increases or decreases in size, when it moves from one group in the classification (0, 1–5, 5–9, 10–19 or more than 20 workers) to another. One disadvantage of using the DIRCE database is that we only observe groups of firms by industry and ignore a firm’s individual trajectory. There have been various formulations of the dependent variable used in the literature, such as the arithmetic mean rate of growth, the annual rate of growth or the difference in size. However, Pesaran and Smith (1995) point out that fixed effects can increase the problems of heterogeneity in b, and propose an indirect estimation of a and b with an observable variable. We only work with data covering the period 1994–2002 because of data gaps in the sample. We omit firms after the first instance of missing data since we find evidence that missing data is randomly distributed across firms. Applying the alternative of imputations could result in serious measurement error. Delmar and Davidsson (1998) estimate a model for high-firm growth. Thus, our results are different because those authors focus only on high-growing firms.
44
SMEs in a globalised world
REFERENCES Acs, Z.J. and D. Audretsch (1989), ‘Births and firm size’, Southern Economic Journal, 56, 467–75. Arauzo, J.M. and A. Segarra (2005), ‘The determinants of entry are not independent of start-up size: some evidence from Spanish manufacturing’, Review of Industrial Organization, 27, 147–65. Audretsch, D. (2002), ‘The dynamic role of small firms: evidence from the US’, Small Business Economics, 18, 13–40. Audretsch, D. and T. Mahmood (1995), ‘New firm survival: new results using a hazard function’, Review of Economics and Statistics, 77, 97–103. Audretsch, D. and E.E. Lehman (2005), ‘Mansfield’s missing link: the impact of knowledge spillovers on firm growth’, Journal of Technology Transfer, 30, 207–10. Audretsch, D., L. Klomp and R. Thurik (1999a), ‘Do services differ from manufacturing?’, in David Audretsch and Roy Thurik (eds), Innovation, Industry Evolution and Employment, Cambridge: Cambridge University Press, pp. 230–52. Audretsch, D., E. Santarelli and M. Vivarelli (1999b), ‘Start-up size and industrial dynamics: some evidence from Italian manufacturing’, International Journal of Industrial Organization, 17, 965–83. Audretsch, D., L. Klomp, E. Santarelli and A.R. Thurik (2004), ‘Gibrat’s Law: are the services different?’, Review of Industrial Organization, 24, 301–24. Callejon, M. and A. Segarra (1999), ‘Business dynamics and efficiency in industries and regions’, Small Business Economics, 13, 253–71. Calof, J.L. (1994), ‘The relationship between firm size and export behavior revisited’, Journal of International Business Studies, 25, 367–87. Calvo, J.L. (2006), ‘Testing Gibrat’s Law for small, young and innovating firms’, Small Business Economics, 26, 117–23. Carre, M.A. and A.R. Thurik (1998), ‘Small firms and economic growth in Europe’, Atlantic Economic Journal, 26, 137–46. Caves, R.E. (1998), ‘Industrial organization and new findings on the turnover and mobility of firms’, Journal of Economic Literature, 36, 1947–83. Chen, J.R. and W.C. Lu (2003), ‘Panel unit root tests of firm size and its growth’, Applied Economics Letters, 10, 343–5. Commercial Performance Information Bureau of the Bank of Spain [Central de Balances del Banco de España] (1997), Resultados anuales de las empresas no financieras, 1996, Madrid: Servicio de Publicaciones del Banco de España. Correa, A., M. Acosta, A.L. González and U. Medina (2003), ‘Size, age and activity sector on the growth of small and medium firm size’, Small Business Economics, 21, 289–307. Delmar, Frédéric, and Per Davidsson (1998), ‘A taxonomy of high-growth firms’, in Paul D. Reynolds, William D. Bygrave, Nancy M. Carter, Sophie Manigart, Colin M. Mason, G. Dale Meyer and Kelly G. Shaver (eds), Frontiers of Entrepreneurship Research, Wellesley, MA: Arthur M. Blank Center for Entrepreneurship, Babson College, pp. 399–413. Encuesta Sobre Estrategias Empresariales (various years), Las empresas industriales, an annual survey of the activity and business strategies of Spanish manufacturing firms, Madrid: Spanish Ministry of Science and Technology (formerly Ministry of Industry and Energy).
Why do SMEs grow?
45
Ericson, R. and A. Pakes (1995), ‘Markov-perfect industry dynamics: a framework for empirical work’, Review of Economic Studies, 62, 53–82. Fariñas, J. and L. Moreno (2000), ‘Firm’s growth, size and age: a nonparametric approach’, Review of Industrial Organization, 17, 249–65. Geroski, P.A. (1995), ‘What do we know about entry?’, International Journal of Industrial Organization, 13, 421–40. Geroski, P. (1999), ‘The growth of firms in theory and in practice’, Centre For Economic Policy Research discussion paper no. 2092. Gibrat, R. (1931), Les Inegalités Économiqués, Paris: Sirey. Goddard, J., D. McMillan and J.O.S. Wilson (2006), ‘Do firm sizes and profit rates converge? Evidence on Gibrat’s Law and the persistence of profits in the long run’, Applied Economics, 38, 267–78. Görg, H., E. Strobl and F. Ruane (2000), ‘Determinants of firm start-up size: an application of quantile regression for Ireland’, Small Business Economics, 14, 211–22. Hart, P. (1962), ‘The size and growth of firms’, Economica, 29, 29–39. Ijiri, Y. and J.A. Simon (1977), Skew Distributions and the Sizes of Business Firms, Amsterdam: North Holland. Jovanovic, B. (1982), ‘Selection and evolution of industry’, Econometrica, 50, 649–70. Lazarova, S., G. Urga and F. Walters (2003), ‘Are differences in firm size transitory or permanent?’, Journal of Applied Econometrics, 18, 47–59. Lotti, F., E. Santarelli and M. Vivarelli (2003), ‘Does Gibrat’s Law hold among young, small firms?’, Journal of Evolutionary Economics, 13, 213–35. Machado, J.A.F. and J. Mata (2000), ‘Box-Cox quantile regression and the distribution of firm sizes’, Journal of Applied Econometrics, 15, 253–74. Mansfield, E. (1962), ‘Entry, Gibrat’s Law, innovation, and the growth of firms’, The American Economic Review, 57, 1023–51. Nooteboom, B. (1994), ‘Innovation and diffusion in small firms: theory and evidence’, Small Business Economics, 6, 327–47. Oliveira, B. and A. Fortunato (2006), ‘Testing Gibrat’s Law: empirical evidence from a panel of Portuguese manufacturing firms’, International Journal of the Economics of Business, 13, 65–81. Pakes, A. and R. Ericson (1998), ‘Empirical implications of alternative models of firm dynamics’, Journal of Economic Theory, 79, 1–45. Pesaran, H. and R. Smith (1995), ‘Estimating long-run relationships from dynamic heterogeneous panels’, Journal of Econometrics, 68, 79–113. Pfaffermayr, M. (2004), ‘Export orientation, foreign affiliates, and the growth of Austrian manufacturing firms’, Journal of Economic Behavior and Organization, 54, 411–23. Prais, S.J. (1976), The Evolution of Giant Firms in Britain, London: Cambridge University Press. Resende, M. (2004), ‘Gibrat’s Law and the growth of cities in Brazil: a panel data investigation’, Urban Studies, 41, 1537–49. Robbins, D.K., L.J. Pantuosco, D.F. Parker and B.K. Fuller (2000), ‘An empirical assessment of the contribution of small business employment to US state economic performance’, Small Business Economics, 15, 293–302. Segarra, A. and M. Callejón (2002), ‘New firms’ survival and market turbulence: new evidence from Spain’, Review of Industrial Organization, 20, 1–14. Segarra, A., J.M. Arauzo, N. Gras, M. Manjón, F. Mañé, M. Teruel and B.
46
SMEs in a globalised world
Theilen (2002), La Creación y la supervivencia de las empresas industriales, Madrid: Civitas. Sistema de Análisis de Balances Ibéricos (SABI) (n.d.), accessed at www.informa. es. Storey, D.J. (1994), Understanding the Small Business Sector, London: Routledge. Sutton, J. (1997), ‘Gibrat’s legacy’, Journal of Economic Literature, 35, 40–59. Voulgaris, F., D. Asteriou and D. Agiomirgianakes (2003), ‘The determinants of small firm growth in the Greek manufacturing sector’, Journal of Economic Integration, 18, 817–36.
3.
Access of small firms to knowledge networks as a determinant of local economic development Miren Larrea, Alazne Mujika and Mari Jose Aranguren
INTRODUCTION Relocation movements in the global context and related local economic processes (Krugman, 1991, 1998; Fujita et al., 2001) have lately raised the interest of researchers and policymakers in Spain. As Myro et al. (2004) and Fernández-Otheo et al. (2007) illustrate, the manufacturing sector in this country – especially the technologically advanced industries – has been affected by disinvestments in the last decade and inward foreign direct investment (FDI) flows have fallen dramatically. While this trend is seen as a threat by most of the policymakers due to the subsequent rise in unemployment rates, it is considered by others as an opportunity to reinforce the knowledge-based economy. Anyway, important asymmetries can be detected in the process. While some firms are positively reacting to the new context, a significant pool of small firms with a strong local focus are losing their traditional markets without being capable of finding potential new ones. Regarding policies delivered in this context, attention has moved from macro to micro policies, through which advantages of clustering and networking are sought. In this sense, national economic performance is explained not just as the result of macro policies, but also as the result of systemic interactions among different local processes. This focuses attention on the regions of Spain, where different policies to help SMEs are being implemented. In this case, the Basque Country is analysed, where industrial policy is mainly a competence of the Basque Government. But clustering and networking dynamics do not always respond to the regional units of analysis either. In this case, a more local (subregional) geographical unit, the county, is studied. This responds to the hypothesis that the same way regions play a role to understand competitiveness of 47
48
SMEs in a globalised world
nations, smaller units of analysis are also needed to understand economic development of regions. The second hypothesis behind the research project is that there is a significant part of the firms with between 10 and 49 employees in the Basque Country, which are confronting globalisation in isolation due to a significant extent to size-related factors. Up to now no policymakers nor firms themselves have found formulas to break such isolation and this is seen as a problem not only for small firms themselves, but for the local development of areas where these firms have a significant role. To reach conclusions for policymakers that could help overcome this situation, this chapter focuses on a series of key issues that could help small firms improve competitiveness through networking: the concept of competitiveness itself, innovation capacity, knowledge management, management models and cooperation networks. The field study has been developed in the Middle Urola Valley (Basque Country, Spain) where specific policy measures have followed the research project. In this valley of approximately 25 000 inhabitants a network of firms, training centres and local councils was constituted by the development agency, Iraurgi Lantzen, to foster collective learning and innovation in management. The name of the network is Ezagutza Gunea. Experience shows that the network has proved to be effective for those firms with 50 or more employees (Aranguren et al., 2007). But paradoxically, the smallest firms – which we could expect to have more need for a networking structure than bigger ones – cannot participate because they do not have the required structure. That means that these networks, that in theory should provide small firms with mechanisms to overcome their weaknesses due to size, fail to do so because there are minimum size and structure requirements to participate in them. In this context, the research reported in this chapter aims to identify the characteristics of small firms with between 10 and 49 employees that have limited access to knowledge and innovation networks and to make a proposal to related public authorities as to how these small firms in the area could be integrated in this Ezagutza Gunea network. To do so, we first present the main concepts on which the diagnosis is based. In the next section the research methodology is described and in the following section the results are shown. Finally, the main conclusions are presented.
Access to knowledge networks and local economic development
49
KEY ISSUES FOR COMPETITIVENESS IN SMALL FIRMS This section presents the main concepts that confirm the theoretical background for the chapter and have been later used to define a questionnaire. Details about this questionnaire are presented in the research methodology section. Presentation of Key Issues Association as a formula to enhance competitiveness has been considered a relevant issue by policymakers at a regional level. Some of the contributions that have reinforced this idea have been the definition of clusters by Porter (1990) or the contributions by Cooke in the field of regional innovation systems (Cooke and Morgan, 1998; Cooke, 2004, 2007; Cooke and Piccaluga, 2006). But the issue of associations is not valid at the regional level, as can be clearly seen in the literature on industrial districts (Brusco 1982, 1990; Becattini, 1990) where geographic units composed by one or a few municipalities are analysed. For this chapter, it has been assumed that firm competitiveness is the basic factor that will ensure local development. Small firms need to seek competitiveness as bigger ones do and, in such a place as the Basque Country where labour costs are no longer low, this requires innovation capacity. Innovation is mainly based on the persons that configure the firm and their knowledge and interrelations (which go further than R&D). To understand how this knowledge is managed two other factors have been analysed, the management models used by the firms on the one hand and cooperation networks, which might give access to external knowledge and break isolation on the other hand. This way, innovation capacity, knowledge management, management models and cooperation networks are the key issues that have been considered to try to understand how cooperation should help small firms to be more competitive. These key issues have been interpreted considering the characteristics of small firms according to OECD (1995, 1996) and Asheim et al. (2003): ●
●
●
Centralised decisionmaking by the owner/manager, whose features (training, experience, motivation, proactivity) have an important incidence in results. Scarcity of resources, not only of financial ones, but also of human resources, which have an important incidence in management capacity. More flexibility in the short run, though not necessarily in the long
50
SMEs in a globalised world
●
run. Due to its simpler structure, they are able to take short-term decisions faster, but in the long run, their disadvantage in human resources affects negatively in their strategic decisionmaking. Limited strategic vision and weak information systems. Small firms have weaker information systems than bigger ones, which is a handicap for strengths, weaknesses, opportunities and threats (SWOT) analysis and strategic planning.
In the following paragraphs the interpretation made of each of the key issues is presented in more detail. Firm competitiveness The concept of competitiveness can be used at macro, meso or micro level. In this study the micro level concept of firm competitiveness has been used. One of the definitions of firm competitiveness that has been quite broadly accepted is the one given by Salas Fumas (1992) who defines competitiveness as the capacity of firms to continuously grow their market shares keeping their financial viability. In other words, competitiveness is the ability of a firm to provide products and services more effectively and efficiently than relevant competitors. Innovation capacity During recent decades there has been an important change in the way innovation processes are understood. In the 1950s and 1960s innovation was seen as a linear process whereas during the 1980s models based on an interactive view of innovation were developed (see Chaminade and Roberts, 2003). This last point of view is the one adopted in this chapter. This way, innovation is considered as a process of developing mechanisms that will make it possible to create, expand and use all kinds of knowledge and generate collective learning. This is an integrated process (where all areas of the firm participate) and it is network oriented (based on interaction with clients, suppliers and other institutions). Besides, following Lundvall (1992) or Smith (1994), innovation has been understood as a social and technical process and an interactive learning process among firms and their environment. From such a point of view, knowledge is the most important resource, learning the most important process and cooperation a relevant strategy. Regarding firm size, although the linear model clearly favoured big firms, the interactive view opens a debate to whether or not size is as important as it was. Anyway, the disadvantage that small firms have for strategic planning is still an important handicap, because they lack the formal planning processes to support them. Consequently, innovation
Access to knowledge networks and local economic development
51
processes, mainly based on knowledge and experience, are quite informal in small firms. Advantages and disadvantages of small firms related to innovation have been studied (Nooteboom, 1994; Cooke and Wills, 1999; Asheim et al., 2003). According to this last author, the main objectives of innovation processes in small firms are product redesign and cost reduction, and the most common ways to innovate are buying new technology or making small improvements in products to adapt them to client requirements. The main difficulties of innovation that small firms are faced with are the lack of technical know-how, limited human and financial resources, low cooperation and the disadvantages these factors present for strategic planning. Knowledge management The knowledge management part of the questionnaire is based on Nonaka and Takeuchi (1995) and it intended to define how firms manage their knowledge and how this is affecting their innovation capacity. These authors say that if innovation is to be generated, there must be continuous interaction among tacit and explicit knowledge. Nonaka and Takeuchi define four types of processes through which explicit knowledge transforms into tacit and tacit into explicit: externalisation, combination, internalisation and socialisation. According to Hitt (1995), there are different features of the firm which affect those four processes: the leadership style of management, the existence of a strategic plan and the capacity of the persons in the firm to learn and cooperate. Management models In order to understand how knowledge and innovation are managed, the next elements have to be considered besides the scarcity of financial and human resources and short-run flexibility already mentioned: ●
● ● ● ●
Labour conditions (salaries, timetables, security) are worse in small enterprises than in bigger ones and employees are, in general, better prepared in big firms than in small ones (OECD, 1996). Anyway, it is frequently argued that small firm employees are more motivated than those who work for a big firm. Small firms have difficulties investing in new technologies, because of their scarcity in financial and human resources. Due to their size small firms possess less negotiation power than bigger firms. They have a different approach to communication, one that is based more on personal relationships. Small firms have limitations for distribution.
52
SMEs in a globalised world ● ●
Limitations for internationalisation of small firms are due to limited information, experience, qualification and financing. Small firms use less advanced management tools than bigger firms (Aguirre et al., 2003).
Cooperation networks The cooperation part of the project has been defined following Havnes and Hauge (2004) as an interrelation among independent firms or associates that combine their efforts and resources on a value creating process. When such cooperation is established with the goal of creating a stable interrelationship, it has been considered as a network. Finally, social capital has been considered as a crucial element that will make such cooperation and networks work. Social capital has been defined following Nayaran (1999) as the relationships, attitudes and values, networks and norms that configure the quality and quantity of the relationships among people, and contribute to economic and social development. According to Havnes and Hauge (2004), the main reasons SMEs have to cooperate are access to new and bigger markets, to provide a wider product range, access to know-how and technology, additional production capacity, cost reduction, and access to labour force and capital. Most small firms cooperate with a limited number of enterprises and they have mainly informal cooperation. The main obstacles to cooperation that small firms have to contend with are ensuring the preservation of their independence, the lack of information about possible partners, the fear to share information, the implicit risks in cooperation with other firms, and fiscal and legal restrictions. Networking policies require a deep understanding of how small firm managers think and act, so that they will be targeted to those critical issues that are nowadays acting as a restraint for such managers. That is why the key issues defined in this section have been used to design a questionnaire that will give a deep understanding of how small firms compete. Afterwards, the specific items defined to measure such key issues will go through a statistical process to define different groups of firms for which networking policies should have a different focus. The final goal is to make recommendations to local and regional policymakers on the key issues that could catalyse the development of a new competitive culture of small firms through networking policies. Approach for the Empirical Analysis Based on the concepts presented above, the empirical analysis is built on two premises. The first one is that competitiveness of firms is linked to their
Access to knowledge networks and local economic development
53
performance regarding innovation, knowledge management, management models, cooperation and networks. The second premise is that networking policies for SMEs must be closely adapted to their specific context, so that policymakers have a clear understanding of the different patterns followed by firms regarding such issues. Considering that in the Basque Country no previous research has been done in this field, a specific questionnaire has been designed based on these two premises – its development can be seen in the next section. The character of the study is basically exploratory and descriptive. First, the analysis is focused on each variable, due to lack of previous information in this specific field. This description is developed to help researchers and policymakers familiarise themselves with the analysed phenomenon. The results are presented below. Once the basic information related to each variable has been considered by researchers, it will be possible to further focus on the empirical analysis. Due to the policymaking goals of the project, the identification of patterns of behaviour of firms that are leading them to different competitive performances has been prioritised. This could be used by policymakers to adapt their networking policies to each specific group. This goal is reached through descriptive analyses, more specifically, with a factor analyses. The results are presented on page 56.
RESEARCH METHODOLOGY The empirical objective of this research is to know the characteristics of small firms (10 to 49 workers) located in the Middle Urola Valley regarding each of the key issues presented previously and to define groups that have different performance patterns related to such key issues. Research Objectives and Questionnaire Design The next specific objectives for the research have been derived from each of the key issues and specificities of small firms presented in the previous section. The questionnaire design has been based on them. 1.
To diagnose size specific features of small firms in the Middle Urola Valley Attitude towards decision making – proactivity in decisionmaking; importance given to management autonomy; resource availability perception; decisionmaking agility perception.
54
2.
3.
4.
SMEs in a globalised world
To diagnose knowledge and innovation management in the small firms in the Middle Urola Valley Information systems quality and planning; knowledge management; innovation process: definition, sources, strategy, type, drives and breaks, difficulties, contributions. To diagnose management models and other internal features in the small firms of the Middle Urola Valley Finance access perception; workers’ productivity perception; supplier, customer, and competitors’ power; management tools used. To diagnose cooperation and access to networks of small firms in the Middle Urola Valley Cooperation attitude; networking attitude.
Based on these research objectives, a questionnaire of 163 questions was prepared which can be obtained by contacting the authors.1 Before starting the interviews, a pilot questionnaire was undertaken with four firms in order to ascertain that the questionnaire items were correctly understood and that they were not difficult to answer. Sample Design and Coverage The population for this research have been all manufacturing firms located in the Middle Urola Valley with 10–49 employees. The population sample frame was based on internal information supplied by Iraurgi Lantzen, the local development agency, which was contrasted with data from Eustat,2 the Basque Statistics Office (see Table 3.1). Given the population size, it was decided to contact all firms in the population and make personal interviews with the owner/manager. Interviews were held from October 2005 until April 2006. In total, 50 valid answers were obtained (47 per cent of the population). The main industries in the area are well represented in the sample except in the case of the construction sector. This is an especially difficult sector to reach as many of the firms in the population are very small and have very small management structures. Variable Measurement In order to apply the appropriate statistical techniques, first, normality of the variables was contrasted applying the Kolmogorov–Smirnov contrast. This contrast has data adjustment to the normal distribution as the null hypothesis. So a value of associated p higher than 0.05 (signification level of 95 per cent) is required to accept that the data are adjusted to the normal distribution. When data are not adjusted to the normal distribution,
55
73 74 85
52
45 51
36
29
15 18 19 20 22 24 26 27 28
Food industry Textile and clothing Leather and footwear industry Wood and cork industry Paper, publishing and graphic arts Chemical industry Non-metal industry Metallurgy Metal industry, except machinery and equipment Machinery and mechanical equipment construction Furniture industry, other manufacturing industries Construction Wholesale commerce and commerce intermediaries, except motor vehicles Retail commerce, except motor vehicles commerce Investment and development Other management activities Water and food laboratories Total – 2 – 66
1
7 2
12
13
2 1 1 4 1 – 4 1 15
10–20
1 – – 21
–
3 1
4
3
1 – – – – – 1 2 5
21–30
– – – 10
–
3 –
3
3
– – – – – – – 1 –
31–40
– – 1 9
–
1 –
2
4
– – – – – 1 – – –
41–50
1 2 1 106
1
14 3
21
23
3 1 1 4 1 1 5 4 20
Total
Distribution of the population
Distribution of the population and sample
NACE Industry
Table 3.1
0 0 0 24
0
0 1
7
4
0 1 1 1 0 0 0 1 8
10–20
0 1 0 17
0
3 1
3
2
0 0 0 0 1 0 2 0 4
21–30
0 0 0 2
0
0 0
0
2
0 0 0 0 0 0 0 0 0
31–40
0 0 1 7
1
1 0
0
3
0 0 0 0 0 0 0 0 1
41–50
Distribution of the sample
0 1 1 50
1
4 2
10
11
0 1 1 1 1 0 2 1 13
Total
56
SMEs in a globalised world
non-parametric testing is required. This is what happened in the project, as the variables did not adjust to the normal distribution. That is why frequencies were analysed and a factor analysis applied.
EMPIRICAL RESULTS Main Descriptive Results In the next paragraphs, a first approach to the results is made by a descriptive analysis of frequencies. The goal is to highlight some relevant issues that will be used to make policy recommendations in the last section. Size specific features Results on whether the owner/manager is proactive or reactive are not conclusive, as there is evidence in both directions. Regarding the importance given to management autonomy, 75 per cent say that they would accept to lose some autonomy if that was what was needed to make the firm grow, and there are very few who say that the most important thing for them is to maintain such autonomy. Regarding resource availability, 80 per cent of the interviewed owner/ managers say that the human and financial resources are more limited in a small firm than in the medium or large ones. Anyway, only 45 per cent say that the management capability is more limited in the small firm than in the middle or large firm. The flexibility of the small firm to make decisions might, in theory, partially make up for other disadvantages. In this respect, 90 per cent of those owner/managers interviewed say that in their firm the daily decisions are made easily, but the percentage reduces to 45 per cent when the strategic decisions are mentioned. Information systems and knowledge management The majority of the interviewed owner/managers (68 per cent) say that the internal information in their firms is systematised and used to understand the firm and make decisions. But only 54 per cent of them say that they use external information sources to understand the firm and make decisions. Regarding planning, 50 per cent of the interviewed owner/managers have a strategic plan in the firm, but 54 per cent say that the diagnosis and the annual plans are not written down. Anyway, 77 per cent of them consider that the formalisation of this information adds value to the firm. About the way tacit knowledge is explicated and combined with other explicit knowledge, 65 per cent of the interviewed owner/managers say
Access to knowledge networks and local economic development
57
that in their firms they collect technical information about the production processes in written documents. Regarding socialisation, 73 per cent of the owner/managers say that they do transmit the vision to the rest. But only 62 per cent of them think that people in the firm really understand and internalise it. With respect to client knowledge management, 70 per cent of the interviewed firms said that they visit their clients frequently, but only 56 per cent say that they write down what the client says. Anyway, 78 per cent of the interviewed say that all the employees know what the client needs. Another way to share tacit and explicit knowledge is constituted by the meetings related to daily activity. This is also a measure of participation in the firm. In this sense, only 28 per cent of the interviewed say that in their firm the blue collar employees meet to solve problems and improve processes. Innovation process When asked to choose the definition of innovation that best suits them, most of the interviewed owner/managers agree with an interactive view of innovation instead of the lineal model. Specifically, 55 per cent of them say that innovation requires sharing information between everyone in the firm and the external agents. Anyway, 30 per cent of them accept that the innovation is basically a way to satisfy the demand and needs of the market, which does not fit with such an interactive view. Asked about the innovation sources, 57 per cent of the owner/managers say that the main ideas for innovation are developed in the firm and 55 per cent say that they are given by the client (they could choose more than one option, which is why the total sums more than 100 per cent). It is also remarkable that 25 per cent of the interviewed say that only managers participate in the innovation process and 41 per cent say that the managers and the technical office do so. Questions were also posed regarding relations of the firm with the agents of the regional innovation system. About that, 61 per cent of the interviewed said that they do not know any agent of such system and 79 per cent affirm that they have never participated with any of these agents in a project. Regarding innovation strategies, 94 per cent of the interviewed say they do not to have a written strategy, but nearly 60 per cent plan to do some innovation. Management tools used The management tools that are more frequently used are labour risk prevention procedures, quality procedures and continuous improvement systems.
58
SMEs in a globalised world
Cooperation attitude and networking It is noteworthy that 60 per cent of the interviewed owner/managers say that they do not cooperate with other firms. When asked about the reasons for not doing so, 45 per cent say it is because of lack of information about the partners, 34 per cent say it is because they are afraid to show information and 11 per cent of them say that the most important thing for them is to maintain their independence. The main advantages of the cooperation, they perceive, are the possibility to access know-how and technology, to access new and larger markets, the possibility to have more products and to reduce costs. Although 60 per cent of the interviewed say they do not cooperate, 88 per cent affirm that cooperation helps to improve competitiveness. Participation in networks is weak in the cases analysed as only 42 per cent of the interviewed affirm they participate in firm networks (mainly the provincial firm association). When asked about networks with technology centres, universities or a combination of those, only 10 per cent say they participate in one. Definition of Profiles for Policymakers After carrying out the descriptive diagnosis on each of the items derived from the key issues defined, it is necessary to order the data in way that enables conclusions to be drawn and facilitates a policy discussion. One of the first steps was to make a cluster analysis that could define groups of firms with common characteristics. But due to the reduced number of firms analysed, this was not possible. That is why different correlations between the most relevant variables have been analysed to see how they relate to each other and whether significant patterns can be defined. Results show that the variable that has been considered as the best indicator of competitiveness, geographical distribution of the sales, has significant correlations with most of the items that are supposed to affect competitiveness. To complete such correlation analysis, that relates pairs of variables, a factor analysis has been developed. This analysis relates all variables, so a more systemic view is obtained. This adds value to policymakers that need to understand quite a complex context for networking. In the following paragraphs the selection of the geographical distribution of the sales as a central item is justified and the results of the factor analysis presented. Exports have frequently been used as indicators of competitiveness, arguing that a firm that is capable of selling abroad shows the ability to compete internationally. Competitiveness has been established as the final goal to which the rest of the analysed factors should contribute. That is why the variable ‘geographical distributions of the sales’ has been selected as a central one for the argumentation of the conclusions.
Access to knowledge networks and local economic development
59
Factor 2
0.8
0.4
Gipuzkoa
Outsourcing
Access to new technology Found new markets Europe
Spain
0 Invest even in debt Know environment–adapt to environment
Local
–0.4 Direction autonomy importance
Basque country
–0.8
–0.8
–0.4
0
0.4
0.8 Factor 1
Figure 3.1
Factor analysis of attitude towards decisionmaking; importance given to management autonomy and geographical distribution of the sales
Regarding distribution of their sales, firms have been asked for the percentage of sales in the county where they are located (Middle Urola Valley, 25 000 inhabitants); in their province (Gipuzkoa, 600 000 inhabitants); in their region (Basque Country, 2 800 000 inhabitants), in Spain, Europe, USA, Asia and others. To analyse how the geographical distribution of the sales is related to the rest of the items, three factor analyses have been defined. When interpreting the figures, it is important to note that some variables have been reversed. 1. 2. 3.
Attitude toward decisionmaking. Importance given to management autonomy. Geographical distribution of the sales.
The analysis has been done with the statistic program SPAD. The first factor explains 21.81 per cent of the variance and the second one explains
60
SMEs in a globalised world
15.12 per cent. In total, the two factors explain 36.96 per cent of the variance. This first analysis distinguishes on the one hand firms that sell in what can be called the local market (county – measured by the variable ‘local’; and province – measured by the variable ‘Gipuzkoa’) and those that sell further away in non-local markets (measured by the variables ‘Europe’, ‘Spain’ and ‘Basque Country’). Using the capacity to sell away from local markets as an indicator of competitiveness, the first group could be defined as less competitive than the second one. Firms selling in the local market are more reactive in decisionmaking terms (measured by the variable ‘know environment–adapt to environment’). They only worry about finding new markets when they have excess production, rather than actively seeking new markets, and they try not to outsource (measured by the variables ‘found new markets’ only when excess of stock and outsourcing). Before buying a new technology, they wait until it is tested in the market (measured by the variable ‘access to new technology’). Besides, they prefer to maintain their autonomy in decisionmaking instead of growing (measured by the variable ‘direction autonomy importance’). Information systems quality – geographical distribution of the sales In Figure 3.2, the first factor explains 16.00 per cent of the variance and the second one explains 13.75 per cent, so, the total variance explained with those two factors is 29.75 per cent. Firms that sell in the local market have lower quality information systems than those firms that sell in the non-local market. More specifically, local market-oriented firms have not systematised internal information systems (measured by the variable ‘internal information not systematised’). They do not have written technical information (measured by the variable ‘technical information about processes in written documents’). They use mostly internal information for decisionmaking (measured by the variable ‘making decision with internal information’). Besides, they do not have written documents containing an actualised view of their strengths, weaknesses, opportunities and threats or an annual action plan (measured by the variable ‘diagnosis and annual plan not written’). They do not have a strategic plan either (measured by the variable ‘strategic plan done’) and they consider formalisation of the information in written documents as a loss of time (measured by the variable ‘formalisation of the information losing time’). Finally, they do not periodically visit their clients nor write down their indications (measured by the variables ‘visit frequently the client’ and ‘collect the written information given by the client’).
Access to knowledge networks and local economic development
61
Factor 2
0.8 Spain
0.4
Making decision with internal information Internal information not systematised
Europe White collar met with blue collar to improve processes White collar met to improve processes Technical information about processes in written documents Collect the written information given by the client Basque Country
–0.4
Strategic plan done Formalisation of the information losing time Gipuzkoa Diagnosis and annual plan not written Local
Visit frequently the client
Employees see their work contribute to the aim of the firm
Employees with less experience work closely with those experienced Managers say clearly where the firm must go Employees know what the client needs
–0.8
–0.8
–0.4
0
0.4
0.8 Factor 1
Figure 3.2
Factor analysis of information systems quality and geographical distribution of the sales
Innovation process – geographical distribution of the sales In Figure 3.3 the first factor explains 22.10 per cent of the variance and the second one explains 13.75 per cent. The total variance explained with both of them is 36.14 per cent. Firms that sell in the non-local market are more active in the innovation processes. More specifically, firms selling further than the province know better the agents of the regional innovation system (measured by the variable ‘know the agents of the regional innovation system’). They have participated in a project with such agents at least once (measured by the variable ‘participate with a regional innovation system agent’). They have a written innovation strategy (measured by the variable ‘innovation strategy written’) and they expect to do an innovation in the short term (measured by the variable ‘expect to do an innovation’). In the last 10 years they have made at least one innovation in new products (measured by the variable ‘innovation in new products done’), or new ways of organisation (measured by the variable ‘innovation in new ways of organisation
62
SMEs in a globalised world Factor 2
0.8
Know the agents of the regional innovation system Basque country
Innovation in new products bought Participate with a regional innovation system agent
Local Innovation strategy written
n
Expect to do an innovation Innovation in new products done Gipuzkoa
Europe
–0.4
Innovation in new ways of organisation bought
Spain
Innovation in new ways of organisation done
–0.8
–0.8
–0.4
0
0.4
0.8 Factor 1
Figure 3.3
Factor analysis of innovation process and geographical distribution of the sales
done’). Finally, they have bought at least one innovation in new products (measured by the variable ‘innovation in new products bought’), or new ways of organisation (measured by the variable ‘innovation in new ways of organisation bought’). A fourth factor analysis has been developed relating cooperation and networking to the geographical distribution of the sales, but no noteworthy differences have been found between firms that sell in one destination or another. The information obtained with the factor analysis has been synthesised in Table 3.2. Two groups of firms have been detected in the Middle Urola Valley. The first one, which could be catalogued as local market-oriented firms, describes a group of firms that are less competitive, less active on innovation and have less structured knowledge management activities than the second group, the non-local market-oriented firms. Regarding cooperation and networking, no significant differences have been detected between both groups.
Access to knowledge networks and local economic development
Table 3.2
63
Groups of firms derived from the factor analysis Local marketoriented firms
Competitiveness Innovation Knowledge management Cooperation and networking
Non-local marketoriented firms
Lower Higher Less active More active Less structured More structured No significant differences found
CONCLUSIONS The overall aim of this chapter was to identify the characteristics of small firms with between 10 and 49 employees in the Middle Urola Valley that had limited access to knowledge and innovation networks and make a proposal to the related public authorities as to how these small firms in the area could be integrated in the Ezagutza Gunea network. A descriptive analysis has been made to diagnose the situation of these firms regarding each item considered relevant for competitiveness. More specifically, two groups of firms have been defined. The local market oriented one and the non-local market-oriented one. The local market oriented firms represent the more vulnerable group. They are selling mainly in the county and province. Some of them have final products and understand this market as their natural one, but many others were created as suppliers of bigger firms in the area and are now suffering from relocation or import decisions of such big firms. These local market oriented firms have no clear strategic view, have difficulties in understanding their business environment and are very dependent on the mentioned local clients. They constitute the group that would benefit more orientation from local authorities but, at the same time, the more difficult group to access. This is because they are very internally oriented and, like the rest of the firms in the area, have very little cooperative and networking culture. Networking activities developed for this group should be oriented to facilitate information about their business environment and develop competencies to understand such an environment and define a strategy for the firm. The non-local market group has a better competitive position compared with the local oriented one, but it must be considered that although they have access to more distant markets, they still sell very little outside Spain. Anyway, they have more strategic plans and a more active attitude towards innovation. Still, the culture to use cooperation and networks as a tool to access new knowledge is very weak. Networking activities for this
64
SMEs in a globalised world
group could be oriented to improve competencies for internationalisation and innovation. Once the conclusions presented in the previous paragraphs where analysed by local policymakers in the context of the Ezagutza Gunea network, a new member technician was incorporated to the team that coordinated this network with the goal of fostering the participation of firms with between 10 to 49 employees. As a first step, three workshops were held from March 2007 to May 2007 where 28 out of the 50 firms that answered the questionnaire discussed the results of the research and defined possible strategies for cooperation. As a consequence, a plan was defined by 15 of these firms who are now actively cooperating amongst themselves and also with Euskalit, the Basque Foundation for Quality. On the one hand, internal management tools are being improved based on the principles of total quality management. This will make it easier to see short-run results of cooperation. On the other hand, they are working on the development of competences for strategic planning. These results will be seen in the long run, but directly respond to one of the weaknesses detected in the research project, that of the lack of strategic vision of firms based on local markets. Although it is too early to evaluate the impact of the defined policy, the initial goal of using the research results to incorporate firms with between 10 and 49 employees in the area to the Ezagutza Gunea network has been achieved. No doubt this networking activity will help these firms (14 per cent of total population of firms) position themselves in relation to the relocation movements that are affecting them in the global context. This could happen either by improving cost efficiency or by improving their innovation capacity. Only in this last case can we envisage a reinforcement of the knowledge-based economy taking place.
NOTES 1. At the time of publication, authors’ contacts are: Mirren Larrea, larrea@orkestra. deusto.es; Mari Jose Aranguren,
[email protected]; and Alazne Mujika,
[email protected]. 2. See www.eustat.es.
REFERENCES Aguirre, M.S., E. Albizu, J. Chaterina, F.J. Forcada and J. Landeta (2004), Análisis de la calidad en la gestión de las pymes de la CAPV y de su incidencia en los resultados económicos, Bilbao: CONFEBASK.
Access to knowledge networks and local economic development
65
Aranguren, M.J., M. Larrea, A. Mujika, A. Plazaola and R. Triguero (2007), Las empresas pequeñas del Urola Medio. Bases para la competitividad. Azkoitia, Spain: Iraurgi Lantzen. Asheim, B.T., A. Isaksen, C. Nauwelers and F. Todling (2003), Regional Innovation Policy for Small-Medium Enterprises, Cheltenham, UK and Northampton, MA, USA: Edward Elgar. Becattini, G. (1990), ‘The Marshallian industrial district as a socio-economic notion’, in F. Pyke , G. Becattini and W. Sengenberger (eds), Industrial Districts and Interfirm Cooperation in Italy, Geneva: International Institute for Labour Studies, pp. 37–51. Brusco, S. (1982), ‘The Emilian model: productive decentralisation and social integration’, Cambridge Journal of Economics, 6 (2), 167–84. Brusco, S. (1990), ‘The idea of industrial districts: its genesis’, in F. Pyke, G. Becattini and W. Sengenberger (eds), Industrial Districts and Interfirm Cooperation in Italy, Geneva: International Institute for Labour Studies, pp. 10–19. Chaminade, C. and H. Roberts (2003), ‘Fostering innovation in SMEs: when external networks matter. The European experience’, presentation to the 7th International Conference on Technology Policy and Innovation, Monterrey, Mexico. Cooke, P. (2004), ‘Introduction: origins of the concept’, in P. Cooke, M. Heidenreich and H. Braczyck (eds), Regional Innovation Systems: The Role of Governance in a Globalized World, London: Routledge, pp. 2–28. Cooke, P. (2007), ‘To construct regional advantage from innovation systems first build policy platforms’, European Planning Studies, 15 (2), 179–94. Cooke, P. and Morgan, K. (1998), The Associational Economy: Firms, Regions and Innovation, Oxford: Oxford University Press. Cooke, P. and D. Wills (1999), ‘Small firms, social capital and the enhancement of business performance through innovation programs’, Small Business Economics, 13 (3), 219–34. Cooke, P. and A. Piccaluga (2006), Regional Development in the Knowledge Economy, London and New York: Routledge. Fernández-Otheo, C.M., L. Labrador and R. Myro (2007), ‘Deslocalización de empresas y actividades productivas en España: una primera aproximación’, Mediterráneo Económico, 11, 57–78. Fujita, M., P. Krugman and A.J. Venables (2001), The Spatial Economy: Cities, Regions and International Trade, Cambridge, MA: MIT Press. Havnes, E. and P.A. Hauge (2004), 2003 Observatory of European SMEs, SMEs and Cooperation, Luxembourg: Office for Official Publications of the European Communities. Hitt, W.D. (1995), ‘The learning organization: some reflections on organizational renewal’, Leadership and Organization Development Journal, 16 (18), 17–25. Krugman, P. (1991), ‘Increasing returns and economic geography’, Journal of Political Economy, 99 (3), 483–99. Krugman, P. (1998), ‘What’s new about the new economic geography’, Oxford Review of Economic Policy, 14 (2), 7–17. Lundvall, B-A. (1992), National Systems of Innovation: Towards a Theory of Innovation and Interactive Learning, London: Pinter. Mason, C. (1998), ‘La financiación y las pequeñas y medianas empresas’, in Desarrollo y Gestión de Pymes: Aportes para un debate necesario, Los Polvorines, Argentina: Universidad Nacional de General Sarmiento, Chapter 8.
66
SMEs in a globalised world
Myro, R., C.M. Fernández-Otheo and D. Martín (2004), ‘Desinversión y deslocalización de capital extranjero en España’, Ekonomiaz, 55 (1), 106–29. Nayaran, D. (1999), Bonds and Bridges: Social Capital and Poverty, Washington, DC: Poverty Group, PREM, World Bank. Nonaka, I. and H. Takeuchi (1995), The Knowledge-Creating Company: How Japanese Companies Create the Dynamics of Innovation, New York: Oxford University Press. Nooteboom, B. (1994), ‘Strengths and weaknesses of small firms’, Small Business Economics, 6 (5), 327–47. Organisation for Economic Co-operation and Development (OECD) (1995), Las pequeñas y medianas empresas – tecnología y competitividad, Paris: OECD. OECD (1996), SMEs – Employment, Innovation and Growth – The Washington Workshop, Paris: OECD. Porter, M.E. (1990), The Competitive Advantage of Nations, New York: Free Press. Salas Fumas, V. (1992), ‘Aspectos micro-organizativos de la competitividad’, Fundación Empresa Pública working paper no. 9205, Madrid. Smith, K. (1994), ‘Interactions in knowledge systems: foundations, policy implications and empirical methods’, Oslo: STEP Group Report.
4.
Innovation behaviour of Spanish fashion manufacturing SMEs José L. Calvo and Angel L. Culebras de Mesa
INTRODUCTION The Spanish fashion manufacturing industry (SFMI), which includes textiles and clothing, but also fur, leather and shoes,1 has been involved in a very difficult restructuring process. During the 1998–2006 period, the SFMI lost almost 100 000 jobs; its industrial production decreased markedly and its chronic commercial deficit almost quadrupled. This alarming situation is the consequence of trends followed by the fashion business, not only in Spain, but all over the world since the early 1990s. The changes affecting Spain and the European Union (EU) as a whole are mainly related to a number of factors: the globalization/ delocalization process; the size and cost competitiveness of the fashion industry in the new EU members; and full implementation of the Agreement on Textiles and Clothing (ATC), implying the end of the quota system and the complete opening of European fashion markets to tough foreign competitors, in particular from China.2 European fashion industries have reacted to these challenges in different ways, depending on the country. So, Italy, with a special trend since its fashion industry has remained stable and based in an industrial organization model where small firms are locally-integrated in industrial districts, has focused on high value added products supported by magnificent fashion branding (Guercini, 2004); the Netherlands has changed from a production-led to a design-led sector, organizing the design and distribution process while subcontracting or outsourcing production (Scheffer and Duineveld, 2004); in the UK there is evidence of clustering, with a cluster consisting in ‘a group of companies which carry out the functions related to the creation of value in the supply chain, which ends with the production of a garment’ (Jones and Hayes, 2004, p. 274); and Germany’s industry which began a secular decline in the 1980s following outsourcing, has retained higher value-added jobs domestically focusing on ‘the remaining markets in which the additional outlays in production for technological 67
68
SMEs in a globalised world
improvements (textile industry) or quality and fashion (clothing industry) were still rewarded’ (Adler, 2004, p. 306). Far from being protectionist, some government policies, particularly EU policies, are being developed to encourage people to leave the industry (Taplin and Winterton, 2004). This has also been the strategy followed by Spanish authorities since no national public or private measures have been implemented to confront the new international fashion environment. In fact, the public policies applied were directed to reduce the impact of the necessary restructuring process, trying to reaccommodate workers or subsidizing their pre-retirement. At the same time, firms have not tended to restructure into concentrated business entities; and big national groups (Inditex, Mango, Cortefiel) have followed the general international trend, outsourcing production. Many studies have considered solutions to the new situation confronted by the EU fashion industry. EMAP (1997), Observatoire Européen du Textile et de l’Habillement (2000), Stengg (2001), Commission of the European Communities (2003) or Institut Français de la Mode and partners (2004) have proposed measures to the EU, suggesting to ‘foster innovative business models’ since ‘companies . . . need a clear signal that there are successful strategies for the industry based on innovation, creativity and quality, efficient consumer response . . . this reinforces the need to create a real European space for training and research’ (Institut Français de la Mode and partners, 2004, p. 377). In the Spanish case there have also been some proposals about the strategies to be implemented: Montes (2003), Xunta de Galicia (2004), Calvo (2005, 2006) or MITYC (2007) are some examples. In recent years new types of fashion enterprises have emerged (especially in clothing). Most of them have changed from being exclusively producers to becoming a kind of intermediary between customer needs, know-how possessors (including designers) and the potential represented by the global production network. As argued by Adler ‘the producers in industrialized countries can only survive in high level and high technology markets, with and emphasis on know how, specialization and innovation’ (Adler, 2004, p. 303). In this context innovation has emerged as being one of the most important ways to deal with the new competitive environment. Studies concentrating on the relationship between innovation and fashion show the relevance of technology – process and product – and, above all, non-technological innovation – management, logistic, design, marketing, and the use of brands – maintaining the fashion industry in high cost countries. Hines points out a basic modification from traditional supply chains which ‘view flows of goods from upstream raw material suppliers
Innovation behaviour of Spanish fashion SMEs
69
through manufacturing processes and on to the customers’ to the modern supply chain concept ‘viewed from a marketing perspective and as being demand driven by customers. This is a pull-through concept as opposed to a supplier push’ (Hines, 2001, pp. 27–8). Le Pechoux et al. introduce the concept of ‘management innovation’ and define innovation management practices since ‘as products are becoming more varied and complex, creativity rather than productivity is becoming the key to business success and survival’ (Le Pechoux et al., 2001, p. 137). Tungate’s (2005) book is dedicated to the power of brands in fashion markets,3 and logistic is the key variable used to explain the success of Spanish brand Zara (Badía, 2008, Chapter 8). Finally, Culebras de Mesa and Calvo (2007) apply factorial and cluster analysis to explain the behaviour of SFMI and find four innovation patterns. Size and innovation are, probably, the two most important challenges the Spanish fashion industry has to cope with. Size has to be understood first in relation to firms because of the small dimension of Spanish fashion enterprises; but it also refers to the relative importance of the sector in terms of its weight in the Spanish manufacturing industry. This is a very important sector to the Spanish manufacturing structure, especially in some regions (Catalonia and Comunitat Valenciana), where employment reductions would have important political and economic effects. Innovation is also an important challenge since the indicators of research, development and innovation (RDI) place Spain at closing positions in EU15, and this place does not improve if we take only the fashion industry into account. Consequently, the analysis of the innovation activity of Spanish fashion SMEs is a good opportunity to study the possibilities of those firms to surpass a certain difficult future, and, at the same time, to point out measures to improve competitiveness and maintain employment. That is what is proposed in this chapter. Analysing the innovation activities of firms from the SFMI (in short, the fashion industry) by using data from the Spanish Firms’ Technological Innovation Survey for different periods of time. Firms are discriminated by size with the group of SMEs encompassing firms with less than 250 employees, and the other group relating to larger firms. Our conclusions will however concentrate on the SMEs, which form the biggest sub-group of the sample. This division is based on previous studies showing the relevance of the size variable in order to explain differences in terms of the attitude of firms’ innovation in Spain (Calvo, 2000). The results of this study also support this assumption. The conclusions of the analysis seem clear. The size of the Spanish fashion manufacturing industry, basically composed of SMEs, and their innovation attitude are not the best set of instruments to deal with the challenges posed by new EU members and Asian economies. Encouraging
70
SMEs in a globalised world
innovation and creativity in order to focus on value added processes and increasing the capacity to reach foreign markets look like the ways to preserve the SFMI for the future. The chapter is organized as follows: the first section provides information on the European fashion manufacturing industry as a whole, reflecting the relevance of this sector to the European economy as well as the challenges it has to face; the second section is dedicated to the evolution of the SFMI during the last years; this is followed by the main section that presents the innovation attitude of Spanish fashion SMEs compared with big firms; finally, the last section proposes a number of conclusions.
CHALLENGES TO THE EU FASHION INDUSTRY The fashion industry is still very relevant to the economy of the EU27.4 In 2004, the industry was composed of more than 280 000 firms, with a business turnover greater than €243 million and employing more than three million people. This implies that the share of the fashion industry in total European manufacturing value added at factor cost amounts to 4.2 per cent; its share in total manufacturing employment is around 10 per cent; its contribution in terms of the number of firms reaches 12 per cent, and it represents 4 per cent of total European manufacturing business turnover. The industry is concentrated in the most populated countries of the EU, for example, Italy, UK, France, Germany, Spain and Poland. At the same time, some countries, such as Portugal, Italy and Spain can be considered as relatively specialized in this industry, especially in textiles and clothing. The EU25 fashion industry is dominated by micro firms (firms with less than 20 employees) and SMEs, no matter which sub-industry we refer to. Less than 0.5 per cent of firms from NACE 17, 18 or 19 have more than 250 employees, but they represent 23 per cent of employment and more than 26 per cent of value added and business turnover in textiles (NACE 17), and 18 per cent of employment and around 20 per cent of value added and turnover in leather (NACE 20). The analysis of the size distribution shows that the EU T/C sector is predominantly an SME-based industry. Enterprises of less than 50 employees employ 60 per cent of the workforce in the EU clothing sub-sector and produce almost 50 per cent of value added (Commission of the European Communities, 2003, p. 4). More than 80 per cent of the firms are micro firms; their shares in employment and in business turnover and value added at factor costs are more than 20 per cent and 15 per cent respectively. SMEs account for 17 per cent of total firms in NACE 17 and 18, and 22 per cent in NACE 19. But, what is more
Innovation behaviour of Spanish fashion SMEs
71
important is that they represent more than 50 per cent of employment and 60 per cent of value added and business turnover. The EU155 has seen how its fashion industry foundations have been shaken by two major different sources: the new Member States of the enlarged EU and the Agreement on Textiles and Clothing (ATC).6 The fashion industry of the new countries that joined the EU in May 2004 was composed at the time of more than 70 000 enterprises, employing almost one million people. Its biggest producer, Poland, had 300 000 people on the fashion payroll, ranking second only to Italy in that year. Their competitiveness is based on relatively low-wage levels, especially relevant for clothing, the labour intensive segment of the industry. Wage rates in these new Member States ranked from 14 per cent to 42 per cent of the EU15 average wage in 2001.7 The average salary in Poland was 22 per cent of the EU15 average wage in 2001. The situation in Bulgaria and Romania was even worse: those two countries had more than 600 000 fashion employees in 2004, and salaries were 5 and 7 per cent respectively of the EU15 average wage. The second big challenge for the EU15 fashion industry is related to the World Trade Organization Agreement on Textiles and Clothing. In the words of the Institut Français de la Mode and partners ‘the end of the quota system is going to be the beginning of a new phase in the Textiles and Clothing world competition: for every new product with severe quota constraints, many more firms and factories will come onto the battlefield, and more competition, especially on price, fiercer and more ruthless’ (2004, p. 154). This means that European countries will have to increase their competitiveness in order to preserve their market position against competitors which benefit from much lower labour costs and much less constraining regulatory environments. The most special case in point is China, a country characterized by an oversupply of labour; employment in the Chinese textile and clothing sector ranks between 7.5 million employees, according to official governmental data, and 15 million employees, according to estimates by Institut Français de la Mode and partners (2004). The country boasts an integrated textile and apparel supply chain, allowing price competition to be extremely severe.
THE EVOLUTION OF THE SPANISH FASHION INDUSTRY The Spanish fashion industry has been going through difficult times in recent years. Table 4.1 shows the evolution of the industry along a number of selected variables.
72
Textiles (NACE 17) Clothing and fur (NACE 18) Leather and shoes (NACE 19) Fashion industry NACE 17, 18 and 19)
1999 91.5 95.1 115.6 98.8
1998 100.0 100.0
100.0
100.0
313.068
68 811
94 841 149 416
244.667
56 676
77 336 110 655
96.9
113.2
73.4
81.1
65.2
69.6
2005 66.9 57.7
232.332
53 220
76 248 102 864
62.9
65.1
2006 64.0 58.4
216.038
48 213
69 354 98 471
Employment (number of employees) 2000 2004 2005 2006
Industrial production index 2000 2004 90.6 75.7 92.5 63.6
319.545
70 296
74 662
315.321
94 507 154 742
1999
88 929 151 730
1998
−37.1
−34.9
1998/2006 −36.0 −41.6
−99 283
−26 449
−31.5
−35.4
1993/2006 Variation rate −19 575 −22.0 −53 259 −35.1
Evolution of Spanish fashion manufacturing industry: employment, industrial production index and commercial balance of payments
Textiles (NACE 17) Clothing and fur (NACE 18) Leather and shoes (NACE 19) Fashion industry (NACE 17, 18 and 19)
Table 4.1
73 770.2 −1046.7
856.0
−673.7
−1488.9
701.2 −1593.0
666.1 −1523.2
726.1
−2615.6
251.0
−3502.9
−33.8
Fashion commercial balance of payments deficit (million of euros) 1999 2000 2001 2002 2003 2004 −519.5 −583.4 −392.9 −373.3 −502.3 −532.1 −1297.4 −1606.8 −1866.2 −1876.1 −2364.4 −2937.0
−2829.2
−889.8
1998/2004 −80.2 −1859.3
Source: Authors’ calculations derived from Instituto Nacional de Estadística (INE) (various years), Encuesta Industrial de Empresas, Madrid: INE, accessed at www.ine.es/jaxi/menu.do?type=pcaxis&path=%Ft05/p048&file=inebase&L=0.
Textiles (NACE 17) Clothing and fur (NACE 18) Leather and shoes (NACE 19) Fashion industry (NACE 17, 18 and 19)
1998 −451.9 −1077.7
74
SMEs in a globalised world
Fashion employment contracted by 99 283 employees in the period 1998–2006. These years were very severe times for the clothing and fur subindustry, where the number of employees decreased by more than 50 000, and for leather and shoes, with 26 449 jobs lost. In both sub-industries, more than 35 per cent of employees in 1998 had lost their job by 2006. For textile, the reduction was ‘only’ 22 per cent, down by 19 575 employees. The situation is even worse if we take a look at the evolution of the Industrial Production Index. Not one of the fashion sub-industries maintained its 1998 production level in 2006. The fashion industry as a whole8 reduced its production by 37 per cent; the decrease, at more than 40 per cent, was the highest for clothing and fur sub-industry. The Commercial Balance of Payments also deteriorated in the period 1998 to 2004.9 In 2004, the deficit of the fashion industry increased by almost €3000 million, reaching the amount of €3503 million, or 4.2 times the commercial deficit in 1998. By sub-industry, clothing and fur again showed the worst behaviour, since its commercial deficit almost trebled. And, as before, only textile maintained a stable situation increasing its commercial deficit by €80 million. The deficit for leather and shoes increased from €856 million to €890 million over the same period of time. Two characteristics confer the SFMI a specific structure: the geographical and the size distribution of firms. Geographically, the Spanish fashion industry is concentrated in two regions, namely, Catalonia, with 34 per cent of business turnover, and Comunitat Valenciana, with almost 25 per cent of business turnover. At the same time, the Spanish fashion industry is predominantly represented by micro firms and SMEs, with only 0.2 per cent of fashion manufacturing firms having more than 250 employees. By contrast, 92 per cent of these firms have less than 20 workers, leaving the share of SMEs in the total population of fashion firms at 8 per cent. Also, SMEs absorb almost 50 per cent of employment (43 per cent), of business turnover (45 per cent) and of value added at factor costs (48 per cent). Some conclusions can be drawn from the analysis of the Spanish data presented in this section: hard employment reductions began before new challenges were completely in action. Therefore, added competitive forces brought about by the last two enlargements of the EU and by the ATC, have tended to worsen the situation, increasing thereby the adjustment process in the industry. At the same time, the effects of this actual adjustment in the fashion industry will also be concentrated in both small firms and specific geographical areas, highlighting the extent of the problem to policymakers.
Innovation behaviour of Spanish fashion SMEs
75
SPANISH FIRMS’ REACTION TO THE CHALLENGES: THEIR STRATEGY IN TERMS OF INNOVATION As we said in the introduction, innovation has been emphasized as one of the main ways to deal with the new competitive environment in the fashion industry. Therefore, it seems interesting to analyse the innovative behaviour of Spanish fashion SMEs, in order to unveil their ability to respond. This is what is proposed in this section. After a brief analysis of the data and the innovative propensity amongst Spanish fashion firms, the section investigates the characteristics of innovative enterprises and innovation patterns. The data used in this study come from the Spanish Firms’ Technological Innovation Survey (SFTIS) conducted by the Spanish National Statistics Institute.10 The SFTIS gives information about ‘the innovation process structure (R&D and other innovative activities) and shows the relationships between that process and the innovative strategy of firms, the factors that help or obstruct their innovation capacity and the economic return for firms’.11 It is directed to industrial, services and construction firms with more than 10 workers with a sample of 43 000 firms approximately. The survey follows the methodological guidelines defined in the Oslo’s reference book (OECD 2005). The information includes the acquisition of new technologies, the type and number of technological innovations, R&D activities, innovation expenditures and their regionalization, the economic returns and aims of technological innovation, the sources of and obstacles to innovation, and non-technological innovation activities. The information used in this section comes from different time periods. The descriptive analysis employs data from the years 2003–05, except for the leather and shoes sub-industry (NACE 19) for which the reference period is 2001–03. The reason is that data for this sub-industry were not disaggregated by size class due to confidentiality problems in the last period. Therefore when we refer to the fashion industry in Tables 4.2, 4.3 and 4.4 we are talking about NACE 17 and 18 and excluding NACE 19, which has its own row. On the other hand the analysis of innovation strategies and their results, innovation patterns, uses microdata from SFTIS in 2000. That is the last, and unique, year the Spanish National Statistics Institute provides microdata for this survey, which means very complete information about firms’ characteristics and behaviour. Innovation Activities of Spanish Fashion SMEs The first five rows of Table 4.2 present the share of Spanish innovative firms in the respective NACE divisions. As can be seen, only one third of
76
Table 4.2
SMEs in a globalised world
Innovating firms in the period 2003–05 (per cent) SMEs less than 250 employees
Large firms 250 employees and more
Total
43.7 22.4 32.3 23.9 37.6
72.8 43.1 60.0 50.0 80.9
44.1 22.6 32.6 24.1 38.7
Share of firms with non-technological innovations Textiles (NACE 17) 22.3 Clothing and fur (NACE 18) 17.0 Fashion industry (NACE 17 and 18) 19.5 Leather and shoes (NACE 19) (*) 31.5 Manufacturing industry 24.7
43.6 33.5 39.3 37.5 55.5
22.6 17.2 19.7 31.5 25.4
Share of firms with R&D Textiles (NACE 17) Clothing and fur (NACE 18) Fashion industry (NACE 17 and 18) Leather and shoes (NACE 19) (*) Manufacturing industry
56.7 23.9 42.7 50.0 58.8
16.0 2.7 8.9 5.7 12.0
Share of Innovating Firms Textiles (NACE 17) Clothing and fur (NACE 18) Fashion industry (NACE 17 and 18) Leather and shoes (NACE 19) (*) Manufacturing industry
15.4 2.5 8.5 5.5 10.9
Notes: (*) Data from 2001–03. Data disaggregated by size class for the period 2003–05 not available. Source: Authors’ calculations derived from Spanish Firms Technological Innovation Survey 2005, http://www.ine.es/inebmenu/mnu_imasd.htm.
Spanish fashion firms (NACE 17 and 18) innovate. This share is not far away from the percentage obtained for the whole industry, at 37.6 per cent. But what really makes a difference is the decomposition by size: for SMEs, firms with less than 250 employees, the share is the same as for the whole sample; but for large firms with 250 employees and more the fraction of innovators reaches 60 per cent in the textiles and clothing categories, and 50 per cent for the leather and shoes sub-division (NACE 19). Focusing on a sub-industry level shows that the shares are also heterogeneous: the most innovative SMEs are located in the textiles category (NACE 17) (where 44 per cent of its firms innovate) meanwhile the percentages for clothing and fur (NACE 18) and leather and shoes (NACE 19) do not reach 25 per cent. Non-technological innovations12 are only implemented in the Spanish fashion industry by 20 per cent of its SMEs,
Innovation behaviour of Spanish fashion SMEs
77
an amount smaller than the one found on average for the Spanish manufacturing industry, 25 per cent. Nevertheless, the share for the fashion industry increases to 39 per cent if we consider large firms (against 55 per cent for the whole Spanish manufacturing sample). Again, clothing and fur presents the smallest percentages, 17 per cent of SMEs and 33.5 per cent of large firms engage in this type of innovation. By contrast, leather and shoes denotes the biggest shares (31.5 per cent for SMEs and 37.5 per cent for large firms). The last rows of Table 4.2 present the share of fashion firms with R&D activities. The results are clear and denote a very different behaviour depending on size: only 8.5 per cent of fashion SMEs belong to this group meanwhile the average increases to 43 per cent if we consider large firms. The decomposition by sub-industry increases the differences. For textiles, 15 per cent of SMEs innovate against 57 per cent for large firms. In the case of clothing and fur, and leather and shoes, the difference between SMEs and large firms is one to ten: only 2.5 per cent of clothing and fur SMEs and 5.5 per cent of leather and shoes SMEs carry out R&D, against 24 per cent and 50 per cent of large firms, respectively. It is important to note the extremely uncompetitive situation of the clothing and fur sub-industry, with almost 97 per cent of its firms with no R&D activity at all. This descriptive analysis leads to the conclusion that Spanish fashion SMEs innovative behaviour is not adequate in terms of responding to the new competitive challenges mentioned in this chapter. Around 70 per cent of the SMEs did not introduce technological or nontechnological innovations over the period under review, and less than 10 per cent of the SMEs carry out R&D activities. Therefore, we can affirm that Spanish fashion SMEs do not consider innovation as one of the most important ways to improve their competitiveness. Characteristics of Spanish Fashion Innovative SMEs After delineating the trends in terms of innovative strategy amongst Spanish fashion firms, this sub-section focuses on their common characteristics. Table 4.3 analyses some of the most important. The innovating intensity, measured as the share of innovation expenditure on business turnover, is bigger for SMEs than for big firms: in textiles SMEs intensity reaches 2.96 per cent against 1.83 per cent for large firms with 250 employees and more; and in leather and shoes the percentages are 2.43 per cent for SMEs and 1.43 per cent for large firms. Only in clothing and fur the relationship reverses: 2.39 per cent for SMEs against 2.75 per cent for large firms. Table 4.3 also supports the conclusion obtained already from the data observable in Table 4.2 and related to R&D. Even for innovative firms,
78
Table 4.3
SMEs in a globalised world
Characteristics of Spanish fashion innovative SMEs (per cent) SMEs less than 250 employees
Innovation intensity by innovative firms Textiles (NACE 17) Clothing and fur (NACE 18) Fashion industry (NACE 17 and 18) Leather and shoes (NACE 19) (*) Manufacturing industry
2.96 2.39 2.79 2.43
Large firms 250 employees and more 1.83 2.75 2.31 1.43
Total
2.63 2.58 2.61 2.10
Share of R&D expenditures on innovation expenses(**) Textiles (NACE 17) 66.4 Clothing and fur (NACE 18) 32.9 Fashion industry (NACE 17 and 18) 58.0 Leather and shoes (NACE 19) (*) 44.0 Manufacturing industry 52.3
80.4 74.2 76.5 95.7 62.6
69.3 55.8 64.2 55.6 58.8
Share of innovative firms belonging to a group of firms Textiles (NACE 17) 7.5 Clothing and fur (NACE 18) 1.8 Fashion industry (NACE 17 and 18) 5.4 Leather and shoes (NACE 19) (*) 4.1 Manufacturing industry 13.2
50.0 33.3 44.8 100.0 74.3
8.5 2.4 6.2 5.0 16.3
12.1 23.9 17.5 12.5 38.3
10.8 3.3 6.8 2.2 7.2
Share of innovation cooperators Textiles (NACE 17) Clothing and fur (NACE 18) Fashion industry (NACE 17 and 18) Leather and shoes (NACE 19) (*) Manufacturing industry
10.8 3.1 6.7 2.1 6.4
Notes: *(*) Data from 2001–03. Data disaggregated by size class for the period 2003–05 not available. (**) It includes internal and external R&D. Source: Authors’ calculations derived from Instituto Nacional de Estadística (INE) (2005), Encuesta sobre innovación technológica en las empresas, 2000, Madrid: INE, accessed at www.ine.es/inebmenu/mnu_imasd.htm.
innovation is considered from a different point of view depending on firm size: for SMEs, innovation expenditure is concentrated in activities other than R&D, mainly on the purchase of machinery (with the exception of textiles). By contrast, large firms dedicate almost the full amount of their innovation expenditure to R&D. These differences are really important,
Innovation behaviour of Spanish fashion SMEs
79
since they denote a very different way to proxy competitiveness related to innovation. Whereas small firms rely on ‘bought innovation’, large firms develop their own innovating structure. The ‘defensive’ attitude of Spanish fashion SMEs could be related to their ownership and their cooperative approach: SMEs are not integrated in industrial groups and they do not cooperate. This is different for firms with 250 employees and more that tend to belong to groups of enterprises, with an important share of them cooperating with other firms. Nevertheless, the ‘more cooperative propensity’ of large firms is smaller than that of the manufacturing sector on average. Results of Innovation for Spanish Fashion SMEs One of the main arguments to prove the relevance of innovation is to measure the share of business turnover that can be attributed to innovation activity. That is included in Table 4.4, showing the share of new or improved products as a percentage of business turnover. It can be seen that, for SMEs, only 20.1 per cent of fashion business turnover can be explained by new products or by products with noticeable changes. This share is very similar for large firms. The most ‘innovative’ sub-sector is leather and shoes, with shares of new or improved products in turnover ranging from 40 per cent in the case of SMEs to 55 per cent for large firms. Nevertheless, it should be remembered that data for the leather and shoes sub-industry refers to the period 2001–03, meanwhile that for the other two sub-industries refers to the period 2003–05.
Table 4.4
Share of new or improved products in business turnover
Textiles (NACE 17) Clothing and fur (NACE 18) Fashion industry (NACE 17 and 18) Leather and shoes (NACE 19) (*) Manufacturing industry
SMEs less than 250 employees
Large firms 250 employees and more
Total
22.4 14.7 20.1 41.8 24.3
20.5 15.2 19.7 54.7 35.1
22.0 16.6 19.9 44.4 32.2
Note: (*) Data from 2001–03. Data disaggregated by size class for the period 2003–05 not available. Source: Authors’ calculations derived from Spanish Firms Technological Innovation Survey 2005. http://www.ine.es/inebmenu/mnu_imasd.htm.
80
SMEs in a globalised world
Innovation Strategy: Patterns of Innovation in Spanish Fashion Industry In order to analyse the innovation behaviour of Spanish fashion industry firms we use a sample of 909 fashion firms from SFTIS, 257 of them innovative (28.3 per cent).13 The study began with 150 variables and after a rigorous statistical process 87 of them were selected, including all the essential features of innovation. A principal component analysis gave rise to ten factors which explained most of the variance of innovative success of Spanish fashion firms. Factors were grouped using cluster analysis, obtaining four innovation patterns which explain the main characteristics and strategies of Spanish fashion innovative firms. The innovation behaviour of Spanish fashion firms can be classified in the following four patterns, the main characteristics of these patterns are summarized in Table 4.5. Pattern I, the most innovative one, represented by a small number of firms, only 3.5 per cent. These firms concentrate their efforts on external R&D and introduce essentially product innovation. Their results are also very significant, since they have a big share of their business turnover integrated by new products to the firm. They also cooperate with all kinds of institutions and agents of Spanish and European technological systems, which makes a big difference with the rest of the patterns. The way to protect and appropriate innovation is in their use of design and especially brands. Pattern II, the second most innovative group, is made up of 8 per cent of Spanish fashion innovating firms. They have a radical innovative strategy. The firms focus on internal R&D and introduce product and process innovations. But the main characteristic of this pattern relates to the introduction of new products not only for the firm but also for the market. Their lack of cooperation is an important constraint. Pattern III is represented by 15 per cent of firms. They focus on product and non-technological innovations, both adjusted to the necessary dynamic behaviour in the fashion industry. Firms in this pattern concentrate on internal R&D, but their main constraint is cooperation, since they have not found the way to interrelate with other innovative agents. Pattern IV, almost three quarters of Spanish fashion industry firms do not innovate. These firms who do innovate allocate limited funds to this activity. This is the main conclusion we obtain from Pattern IV’s analysis: this group does not cooperate; its innovation protection is marginal, and firms basically introduce process innovation by the way of purchasing equipment with incorporated technology. This is a passive rather than active innovation policy, which will lead these firms to face harsh difficulties in competitive fashion markets in the future.
81
Suppliers Institutions Meetings R&D (internal and external) Equipment acquistion Costs and risks Supply sources Qualified employment Lack of clients’ sensivity
Sources of knowledge
Obstacles to innovation
External R&D Marketing Equipment purchase
Effort concentrated on
Share in pattern (%)
NACE 17: 3.4 NACE 18: 1.7 NACE 19: 5.8 NACE 17: 55.6 NACE 18: 11.1 NACE 19: 33.3
3.5
Pattern I
Costs and risks Supply sources Qualified employment
Costs and risks
Suppliers Equipment purchase
Training Internal R&D Meetings
Costs and risks Qualified employment Lack of clients’ sensitivity
Equipment acquisition
NACE 17: 69.2 NACE 18: 79.7 NACE 19: 75 NACE 17: 54 NACE 18: 25 NACE 19: 21
72.8
Pattern IV
Internal R&D (highest effort) Equipment acquisition
NACE 17: 19.2 NACE 18: 8.5 NACE 19: 11.5 NACE 17: 71.8 NACE 18: 12.8 NACE 19: 15.4
15.2
General features 8.2 NACE 17: 7.5 NACE 18: 10.2 NACE 19: 7.7 NACE 17: 52.4 NACE 18: 28.6 NACE 19: 19.0 Innovative characteristics Internal R&D External R&D Marketing Equipment purchase Meetings Internal R&D Equipment purchase
Pattern III
Pattern II
Main characteristics of Spanish fashion innovation patterns
Size (% in number of firms) Share in sub-industry (%)
Table 4.5
82
Firm new products (31.3%) Market new products (9.2%) Brands Design
Results
Firm new products (24.6%) Market new products (9.9%) Factory secrets, design, leadership
Firm new products (14.9%) Market new products (6.5%) Not relevant
Not relevant (2.7%)
Product innovation (50.3%) Process innovation (67.4%)
Product innovation (84.6%) Process innovation (61.5%) Dynamism. All kinds of non-technological innovations Not relevant (2.6%)
Product innovation (90.5%) Process innovation (76.25) Marketing, new strategies, design 19% Suppliers, EU institutions, Universities, clients Firm new products (28%) Market new products (16.2%) Patents, utility models, brands
Pattern IV
Pattern III
Pattern II
Source: Authors’ calculations. Data from Instituto Nacional de Estadística (INE) (2002), Encuesta sobre innovación technológica en las empresas, 2000, Madrid: INE, accessed at www.ine.es/inebmenu/mnu_imasd.htm.
Appropiation and protection
100% All kind of institutions
Cooperation
Pattern I
Product innovation (77.8%) Process innovation (55.5%)
(continued)
Innovation strategy
Table 4.5
Innovation behaviour of Spanish fashion SMEs
83
CONCLUSIONS: POLICY IMPLICATIONS FOR THE SPANISH FASHION INDUSTRY The fashion industry of the EU15 is faced with a challenge emanating from two different sources: first, the eastward enlargement of the EU, and second the ATC. The results of these challenges have not yet completely affected European countries, but it is almost sure that they are going to intensify the restructuring process that has been affecting the fashion industry during recent decades. This is especially relevant for Spain, where job destruction will affect specific regions and SMEs. All European studies related to the fashion industry have emphasized innovation as an essential way to increase competitiveness and maintain employment. But, as shown in this chapter, the innovation strategies of Spanish fashion SMEs are extremely fragile. This is due to a number of reasons that can be summarized as follows. First, only one third of small and medium-sized Spanish fashion manufacturing firms introduce technological or non-technological innovations, and less than 10 per cent of them engage in R&D activities. Those percentages increase considerably if one considers large firms. Second, the expenditure on innovation by SMEs expenses is concentrated on equipment. By contrast, large firms dedicate their resources mainly to R&D. In fact, many Spanish fashion SMEs innovators do not consider R&D as part of their innovation strategy. Third, cooperation, a theoretical alternative to intra-firm R&D expenditure, is not a real option for most Spanish fashion SMEs. Only 6.7 per cent of these firms cooperate with clients, suppliers, public and private institutions, or other firms. Fourth, less than 6 per cent of Spanish fashion SMEs are integrated in a group of firms. The percentage increases 8 fold, to 45 per cent, if we consider firms of the same industry but with more than 250 employees. Fifth, the only positive result for SMEs is the innovation expenditure intensity, especially in the textile sub-industry. Sixth, only one fourth of the production of SMEs innovators consist of new or improved goods. Finally, most of the innovative activities are concentrated in textiles; meanwhile in the other fashion sub-industries, and especially in clothing and fur, enterprises do not undertake an active innovation strategy. In fact, when we introduce a principal component and cluster analysis to classify the innovative behaviour of Spanish fashion firms we obtain four different innovation patterns: the first one is very small, 3.5 per cent of firms, and it focuses on external R&D and cooperation; the second pattern is also small, 8.2 per cent, with firms’ innovative efforts concentrated in internal R&D and oriented towards radical innovation, introducing new products to the market; the third pattern, represented by 15 per cent of Spanish fashion innovative firms, assigns moderate resources to
84
SMEs in a globalised world
innovation; finally, the last pattern, the biggest one 72.8 per cent, incorporates hardly any innovative firms. The overall conclusions that can be drawn from the data seem clear. The size of firms in the Spanish fashion manufacturing industry, an industry basically composed of non-innovative SMEs – with 98 per cent of firms having less than 250 employees and only 30 per cent of them innovating – combined with their characteristics (such as non-cooperation, their non-belonging to a group of supply technology firms, a low interest and/ or capacity in R&D, based on ‘bought innovation’ and a low rate of nontechnological innovation implementation) and pattern classification, are not the best assets allowing these firms to confront the new competitive environment. It looks as if the severe production and employment adjustments accomplished during the last years are going to continue in the future. Moreover, the geographical concentration of Spanish fashion firms is going to be an additional problem since the employment adjustments are going to greatly affect specific regions (Catalonia and Valencia essentially). Spanish policymakers should therefore find ways to encourage firm’s innovation activities. Studies should be carried out with a view to finding the answers to important questions related to innovation. These questions are for example: why Spanish SMEs do not innovate? Why do they not cooperate? What can be done to increase R&D? Strategies proposed in the past such as ‘Moda de España’ do not seem to have worked, maybe because it presented an ‘artistic’ view of Spanish fashion but it neither implemented economic and innovation policies nor developed a branding strategy to protect and enlarge the industry. Promoting innovation and creativity in order to focus on value added processes; supporting R&D by public and private institutions; increasing cooperation and the capacity to reach foreign markets look like the ways to preserve the Spanish fashion manufacturing industry. At the same time, Spanish fashion enterprises should realize that they can no longer compete on the basis of costs and wages, and that innovation, technological and above all non-technological (branding, marketing, logistics, and so on) are the basis for future competitiveness.
NOTES 1.
2.
In fact the fashion industry considered in this chapter includes sub-industries with NACE codes 17 (manufacture of textiles), 18 (manufacture of wearing apparel; dressing; dyeing of fur), as well as NACE 19 (manufacture of leather and leather products). In some parts of the chapter, NACE 19 is not considered due to missing information. For more on this specific issue, see Institut Français de la Mode and partners (2004) and CEPS (2005).
Innovation behaviour of Spanish fashion SMEs 3. 4. 5. 6.
7. 8. 9. 10. 11. 12. 13.
85
There are many references related to brand’s power in fashion markets. Michael J. Silverstein and Neil Fiske (2003) is an extraordinary example. Centre for European Policy Studies (CEPS) (2005), and Commission of the European Communities (2003). Fashion industry includes NACE 17, 18 and 19. Data from 2004. The EU15 encompasses Belgium, Denmark, Germany, Spain, France, Ireland, Italy, Luxembourg, Netherlands, Austria, Portugal, Finland, Sweden, the United Kingdom and Norway. In 1995 the World Trade Organization (WTO) signed the Agreement on Textile and Clothing, gradually dismantling the Multifibres Agreement of 1973. One of the consequences was that of January 2005 the European protection system based on quotas disappeared, allowing the free entry of products from foreign competitors such as China or India. See http://www.wto.org/english/docs_e/legal_e/16-tex.pdf for the Agreement. Social Security costs are discounted in these calculations. The fashion industry index was built using weighted annual sub-industries indices. The weights come from the share of every sub-industry in the fashion business turnover. The period during which the data is available. Available at: http://www.ine.es/inebmenu/mnu_imasd.htm Available at: http://www.ine.es/daco/daco43/notaite.htm Non-technological innovations include changes in corporate strategy, use of advanced management and brands, and changes in organization, marketing, aesthetic and design. See Culebras de Mesa and Calvo (2007).
REFERENCES Adler, Ulrich (2004), ‘The dominant feature in the economic development of the German textile and clothing industries’, Journal of Fashion Marketing and Management, 8 (3), 300–319. Badía, Enrique (2008), Zara . . . y sus hermanas, Madrid: LID Editorial Empresarial. Calvo, José L. (2000), ‘Una caracterización de la innovación tecnológica en los sectores manufactureros españoles: Algunos datos’, Economía Industrial, 331, 139–50. Calvo, José L. (2005), La Innovación Tecnológica en la Moda: Textil, Confección, Peletería, Calzado y Cuero, accessed at www.uned.es/dpto-analisis-economico1/ finchprof/calvo/articulos/ModaTechnologia1.pdf. Calvo, José L. (2006), ‘Innovation behaviour of Spanish fashion manufacturing industry: size differences’, presentation to the European Network of Industrial Policy (EUNIP) International Conference 2006, 19–22 June, Limerick, Ireland, accessed at www.uned.es/dpto-analisis-economico1/finchprof/jcalvo/articulos/ Irlanda2006III.pdf. Centre for European Policy Studies (CEPS) (2005), ‘The textiles and clothing industry in an enlarged community and the outlook in the candidates states, final report, FIF no. 20030838. Commission of the European Communities (2003), ‘The future of the textiles and clothing sector in the enlarged Europe’, Commission staff working paper SEC (2003) 1345, Brussels. Culebras de Mesa, Luis A. and José L. Calvo (2007), ‘Patterns of innovation in Spanish fashion industry’, presentation to the European Network of Industrial
86
SMEs in a globalised world
Policy (EUNIP) International Conference 2007, 12–14 September, Polo di Prato, Italy. EMAP (eds) (1997), The UK Fashion Report, Market Tracking International Ltd, London: EMAP Fashion. Guercini, Simone (2004), ‘International competitive change and strategic behaviour of Italian textile-apparel firms’, Journal of Fashion Marketing and Management, 8 (3), 320–39. Hines, Tony (2001), ‘From analogue to digital supply chains: implications for fashion marketing’, in Tony Hines and Margaret Bruce (eds), Fashion Marketing, Contemporary Issues, London: Elsevier Butterworth-Heinemann, pp. 26–47. Institut Français de la Mode and partners (2004), Study on the Implication of the 2005 Trade Liberalisation in the Textile and Clothing Sector, consolidated report Paris: Institut Français de la mode. Jones, R.M. and S.G. Hayes (2004), ‘The UK clothing industry. Extinction or evolution?’, Journal of Fashion Marketing and Management, 8 (3), 262–78. Le Pechoux, Beatrice, Trevor J. Little and Cynthia L. Istook (2001), ‘Innovation management in creating new fashions’, in Tony Hines and Margaret Bruce (eds), Fashion Marketing, contemporary Issues, London: Elsevier ButterworthHeinemann, pp. 136–64. Ministerio de Industria, Turismo y Comercio (MITYC) (2007), Informe del Ministerio de Industria, Turismo y Comercio 2006. Sectores y Políticas. Madrid: MITYC, accessed at www.mityc.es/es-ES/IndicadoresyEstadisticas/Informes/ Paginas/Informes HorizontalesEstructurales.aspx. Montes, Vicente (2003), ‘Análisis DAFO de la Moda en España’, in I Ciclo de Jornadas La Moda un Fenómeno Interdisciplinar, Madrid: UNED. Observatoire Européen du Textile et de l’Habillement (OETH) (2000), Key Trends in 2000, Paris: OETH. Organisation for Economic Co-operation and Development (OECD) (2005), Manual de Oslo. Guía para la recogida e interpretación de datos sobre innovación, 3rd edn, Brussels: OECD, European Commission and Eurostat, accessed at www.conacyt.gob.sv/Indicadores%20Sector%20Academio/Manual_de_ Oslo%2005.pdf. Scheffer, Michiel and Marieke Duineveld (2004), ‘Final demise or regeneration? The Dutch case’, Journal of Fashion Marketing and Management, 8 (3), 340–49. Silverstein, Michael J. and Neil Fiske (2003), Trading Up. The New American Luxury, New York: Portfolio, Penguin Group. Stengg, Werner (2001), ‘The textile and clothing industry in the EU. A survey’, European Commission enterprise papers no. 2, Luxemboug: European Commission. Taplin, Ian M. and Jonathan Winterton (2004), ‘The European clothing industry. Meeting the competitive challenge’, Journal of Fashion Marketing and Management, 8 (3), 256–61. Tungate, Mark (2005), Fashion Brands: Branding Style from Armani to Zara, Spanish edn 2008, Barcelona: Gustavo Gili S.L., Great Britain and United States, Kogan Page. Xunta de Galicia (ed.) (2004), Estudio de necesidades formativas del Sector Textil y de la Confección en Galicia. Consejería de Asuntos Sociales, Empleo y Relaciones Laborales, Galicia, accessed at www.traballo.xunta.es/contenidos/ es/menu_vertical/publicaciones_estadisticas/sec_publicaciones/publicaciones/ publicacion20/publicacion_view.
5.
Family-based firms: evidence from the Portuguese furniture and events organisation industries Vitor Braga and Bernadette Andreosso-O’Callaghan
INTRODUCTION Daily business reality cannot be detached from the relationships established between firms, entrepreneurs and other actors involved, such as employees, customers and suppliers. These relationships gain specific characteristics when they are based on familial links. The literature on family firms acknowledges the importance of such relationships and it has produced a considerable number of writings.1 Typically, both at the time of the firms’ creation and during a few years after their coming into existence – a time frame which is the focus of the present study – family firms in Portugal are small- and medium-sized firms. Owing partly to the richness of the relationships existing in family firms, and to the relative youth of the field of family firms (Amann and Jaussaud, 2008), it is a subject where there remains many research opportunities. In the context of Portuguese society as a whole, the family has been playing a decreasing role; therefore, its importance has also diminished in the context of the firm. Social transition in Portugal, a traditional Catholic country lying at the periphery of the European Union (EU), has impacted on the entrepreneurial structure, where both traditional and modern firms (family and non-family firms) cohabit. However, the family still plays a relatively important role in business, with a special incidence in certain industries. This chapter uses two industries – the furniture and the events organisation industries – to characterise the evolution of the changing Portuguese entrepreneurial approach to business. While the furniture industry is a traditional industry, with predominantly family-based firms, the events organisation industry is a modern sector, generally detached from the traditional Catholic values and therefore where the family plays a marginal role. In addition, this chapter analyses the evolution of the traditional family-based firms 87
88
SMEs in a globalised world
regarding management succession; it addresses how these firms have been coping with the modern business challenges such as deeper integration in the EU and globalisation, whilst being constrained by the structure and values of families. In this chapter, two related but different issues are addressed. The first one aims at describing the nature of business when family is at its core, analysing the interrelatedness and, conversely, the potential conflict between both institutions – the firm and the family – and to establish a parallel, with non-family based firms. The second issue explores the mechanisms of family management succession. Although the family may determine the structure of the firm, it may also play a secondary role in business – for example by providing resources. Thus, the family may play a central role in the process of the firm start-up, and, in later stages of the firm development, it may become part of the firm only through participation in its management. Consequently, this chapter is organised around two major sections. The first section explores the contribution of the family to the firm through its participation in the firm’s creation and in finance. The second section analyses a greater involvement level of the family in the firm by studying the issue of management succession, in order to understand the way in which the firm structure obeys a family hierarchy. Beforehand, a brief review of key theoretical perspectives on family firms is proposed. Most of the family firms studied were created after the Portuguese revolution (1974) and have survived the economic (and political) transition of the Portuguese economy. This has permitted an analysis of the evolution of the Portuguese entrepreneurial structure from a peripheral European country that became an EU Member State listed as one of the four ‘cohesion’ countries of the ‘old’ EU to an economy of dynamic firms. To that extent, some reference to the contribution of the EU cohesion funds to the development of the firm are also made. This Portuguese case is ultimately aimed at explaining how relatively young SMEs in one of the peripheral countries of the EU have been able to evolve and to grow in a context of ever deeper EU integration and globalisation. This chapter is a result of in-depth semi-structured interviews with 29 different organisations (including SMEs in both sectors, local governments, and local and national business supporting institutions) conducted between 2005 and 2006. The results presented subsequently emerged from a methodology with elements of grounded theory, in which data and theory were constantly being compared. Periods of partial analysis enabled a refinement of the interview questionnaire, and thus to collect data on the themes that were emerging along the research. The use of two polar case studies (furniture and events organisation industries) enabled a number of
Family firms in the Portuguese furniture and events industries
89
comparisons between two diametrically opposed business realities. While the furniture industry is a traditional manufacturing industry mainly concentrated in two semi-rural areas – with middle-aged entrepreneurs – and entering in the second generation of family-based firms, the events organisation industry is an emerging services industry, concentrated in urban areas (but where the concentration of activities plays a minor role in the region where they locate); the events organisation industry is led by young professionals with an urban attitude, where the family plays a marginal role in the context of business. Throughout this research, informal relationships were identified as key characteristics of business in Portugal. The family plays a more important role in traditional industries and this role has mainly been related to funding and to the start-up phase of the firms.
KEY THEORETICAL PERSPECTIVES ON FAMILY FIRMS The theoretical argument underlying this section is that family-based firms assume different specific features due to the nature and motivation of their ownership and management. For the purpose of this research (due to the nature of the data), a special emphasis is put on the cases where ownership and management correspond to the same individual, and on the cases in which the firm’s management is conducted by the entrepreneur and hence where the succession takes place within the family. Another aspect of the argument is that family succession is determined by the structure of the family, and that differences in the management style arise from succession, creating a potential conflict within the family. A substantial part of the scientific study on family-based firms is undertaken from a strategic management perspective, although other perspectives are also important, such as those from resource-based theory. Strategic management perspectives emphasise the role of firms’ decisions when a family is at the core of the business (Sharma et al., 1997; Chrisman et al., 2003) and also explores the issue of business succession. Resource-based theory considers the family as a strategic resource that firms can use to gain competitive advantages (Chrisman et al., 2003; Zahra, et al., 2004). At the core of this analysis is the definition of family firms. Although there is a lack of consensus here, the definition is particularly important for the study of the demography of family firms and for evaluating their importance in the context of the different economies.2 Watts and Tucker (2001) define a family firm as any enterprise that involves more than one member of the family. However, under this definition, any firm in which,
90
SMEs in a globalised world
for example, any two brothers co-work is considered as a family firm. In a more narrow way, Westhead and Storey (1997) consider a family firm as being a firm in which at least three of four conditions are met: 1. 2. 3. 4.
it has undergone an intergenerational transition; more than 50 per cent of the shareholding is owned by the family; more than 50 per cent of family-members are involved in day-to-day management; and, where the firm speaks of itself as a family firm.
This definition is also problematic, as it narrows excessively the definition of the family firm. For the purpose of this work, a family firm is a firm in which there is a clear involvement of the family, and for which the management is determined within the family, with more than one member of the family involved in the business. The family firms studied here are all small- and medium-sized firms. Arising from any definition of a family firm is the (conflicting) cohabitation of two institutions: the family and the firm; a matter of high importance is which of the two is dominated by the other. Although there is no clear answer to this question, Donnelly argues that ‘the balance between the family interests and the company interests is usually a psychological one, stemming from the family’s own personal sense of responsibility towards the firm’ (1964, p. 94). Despite this conflict, there are advantages from organising production around a family. The family firms’ organisational culture is an important strategic resource, which can be used in its favour. Moreover, when entrepreneurship is embedded in the family itself, it can lead to competitive advantages over other rivals by reducing the cost of operations and making the family firms one of the most efficient organisational forms (Zahra et al., 2004). There are also advantages in terms of knowledge and the identification of opportunities for entrepreneurship. The argument put forward by Zahra et al. (2004) is that the family emphasises the values of the firms, and in the presence of a culture that values the acquisition of new knowledge, the several actors within the firm are constantly exposed to new sources of knowledge. Moreover, in the cases where the founder has been holding the position for a long time, there is a higher likelihood of an inward focus in the organisational culture. This indicates that, within family-based firms, there is an amplification of certain values when they are embraced by the family. In the same way, Gallo et al. (2002) found that whenever the family property is shared with non-family members, there is an increased development of shared goals, such as strategic alliances, as much as an increased capacity to trust the non-family shareholders.
Family firms in the Portuguese furniture and events industries
91
By contrast, other studies highlight the limitations of family firms. In the words of Donnelly ‘too much involvement with the family interest may prevent a company from capitalising on new market developments or on major growth opportunities’ (1964, p. 96). This is due to ‘excessive nepotism’. The concept of nepotism is described by Donnelly (1964) as the advancement of relatives on the basis of family ties rather than of merit. This limits the potential growth of the firm as the opportunities that can be offered to non-family management talent are restricted. Although many of these issues may be common to most business realities, a comprehensive framework of family firms must consider the role of different cultures. In some cultures not only does the family play a more important role, but also, the organisation of the family itself may differ, generating different features on to family firms. Surprisingly, few studies have addressed the importance of culture in the birth and growth of family firms. Of specific note is the work of Howorth and Ali (2001) who acknowledge the importance of culture and ethnicity to family firms’ operation and motivations in Portugal. The concept of family is hence embedded within the set of local culture, being a social institution with different features across different places. To that extent, the manner in which family firms operate also varies according to the culture. In this sense, the organisation of the family, and the social importance of the family assumes different perspectives and thus, the hierarchy of families may also vary. Accordingly, the way business succession occurs may also be of a disruptive nature. The succession in family firms is a critical moment for the firm, depending on the degree of harmony within the family. Howorth and Ali argue that ‘Succession is not always smooth or continuous and can be disrupted by changes in family circumstances or truncated by resistance’ (2001, p. 233). The literature agrees that there is a set of factors that affect the success of succession, as shown in Figure 5.1 The factors affecting business succession are not only related to the successor; the founder also has an important role. In fact, the role of the founder is described as pivotal in determining and sustaining core business values, even after death (Gatrell et al., 2001). Hence, the second generation (and ensuing generations) are, to some extent, always limited by the legacy of the first generation, not only in terms of the market image of the firm, but also through the existence of a set of norms and conventions, created within the firm that characterise the organisation. Therefore, in some way, the succession is always determined by the founder and it generally relies on factors beyond economic rationality. As Howorth and Ali (2001) note, succession selection may be based on cognitive biases, relying on, for example, personal preferences towards certain family members. As a result
92
SMEs in a globalised world
Rewards from the business
Trust in the successor’s abilities and intentions
Willingness to take over
Personal needs alignment
Preparation level of the successor
Family harmony
Relation between owner–manager and successor
Source:
Rewards from the business
Venter et al. (2005).
Figure 5.1
Theoretical model of factors that influence successful succession
the selection of the successor can be based on several criteria: rational business criteria; traditions; emotions; and/or family objectives.
ENTRY AND FINANCIAL SUSTAINABILITY Much of the literature on SME start-ups seems to indicate that the early years of a firm are of major importance for its own continuation. In those early years, the firm is more vulnerable to the external environment and it still lacks a strong organisational capacity to grow internally and to face external challenges. Following this argument, the role of the family and its social network may become particularly important at this stage as a way to overcome these problems. Moreover, as it tries to grow, the firm faces a critical issue in terms of its survival: financing. The first subsection looks at how family cooperative relationships facilitate the firm’s entry into the respective markets. For the two industries selected, different paths have emerged, indicating that the process of business start-up assumes different
Family firms in the Portuguese furniture and events industries
93
mechanisms in each industry. The second subsection will analyse the role of the family in the financing of the firm at the moment of its foundation. Business Start-ups The following aims to describe how the process of start-up occurred in the two industries; the objective is also to highlight the differences between both industries. Using both industries’ insights provides the opportunity to study: (a) how family-based firms may be in a relatively better position to enter the market and; (b) how entry mechanisms are changing in Portugal, given the decreasing participation of the family in Portuguese firms. The codes in the quotes indicate the firms interviewed, and a brief characterisation of these firms is provided in Appendix 1. The furniture industry Most of the firms in the furniture industry, and especially those located in Paços de Ferreira3 are now under a second generation of management, which reflects the fact that in the 1970s a large number of entrepreneurs entered the market. In terms of the firm’s creation, very similar mechanisms were found in most of the firms studied for the purpose of this analysis. It has been common practice in Portugal (and in some cases it still is) to set up a business in a very informal way. Typically, during after-work hours, an entrepreneur would produce some pieces of furniture on his private premises. Very often, production is undertaken on an informal economic basis. The extra gains from the home-made furniture provide the entrepreneur with financial resources to start-up the business formally. Furthermore, this process also enabled a reduction in the risk of formally starting-up a firm in terms of market access. In Portugal, these firms are known as ‘Empresas de vão de escada’ (literally meaning firms that operate under a set of stairs). The expression indicates that these firms operate on an informal basis lacking proper premises, and that they are not registered publicly. Thus the only way to access these firms is through personal contacts, through the social network of the entrepreneur and the family. The nature of these start-ups is well described by interviewee LB (123) when explaining the beginnings of his father’s and uncle’s firm. It started in a ‘barraco’.4 It was my grandparents’ place. My father often says that they had to put the cow outside and they went inside to produce furniture. And then it grew . . .. They moved to different premises [a different ‘barraco’] and then they came back and started building a proper factory, but very slowly. (LB: 123)
94
SMEs in a globalised world
This quote indicates the informality of activities in the early years of the firm. These entrepreneurs are in an advantageous position in relation to other firms; since they do not pay taxes, their costs are lessened and these small family firms benefit therefore from a cost leadership advantage. From a different perspective, organisation costs are equally low, as it is a one-man firm. A very important dimension of the furniture firm start-up process is family involvement. In fact, the family seems to play a major role in establishing the firm, in various ways (for example, borrowing money, providing resources such as premises). In some (although very few) cases, the first generation’s parents’ experience in the business was a determinant to enter the industry, as interviewee AM (7) explains: My father created the firm in 1972. He was following the tradition from both my grandparents (from my father’s family and from my mother’s family), who used to have their own factories. So, my father worked with his father, and then he worked with his father-in-law and then started up his business. (AM: 7)
However, the most common case is where entrepreneurs worked for a large furniture firm (a few of them existed at that time) and through the mechanism of informality they created their own firm. Across the different firms interviewed, a historical track of family connections seems to repeat itself in most firms. As mentioned earlier, apart from the role of parents in setting up a number of favourable conditions for the creation of the firm, such as providing premises or finance, a joint venture of two or more brothers also seems to be a common mechanism of firm creation. The following quotes (from different firms) help understand the role of this mechanism. There were two partners, in 1977 (the foundation year), who were brothers. They had worked together for about 20 years. At that time the firm was split into two firms: one producing and the other one selling. Meanwhile the two brothers separated and one kept the production firm and the other one the commercial firm. (PP: 7) The firm started up when my father came from the colonial war [early 1970s] with his brother. They then separated and his brother now owns a pool table factory. (PJ: 26) In the beginning there were three partners, three brothers, and then they went to the colonial war [not necessarily together]. After some time, arguments started to appear and they decided to set up businesses individually. My father kept the original firm because he is the eldest. (LB: 130)
The quotes indicate that the partnership between brothers is one of the ways to face the financial demands of setting up a firm. Once the first years
Family firms in the Portuguese furniture and events industries
95
are over, a need for independence seems to overtake financial needs, and then the brothers split up and create individual firms. The third quote also embodies a very important issue in terms of family business – the existence of a family hierarchy. The uniqueness of the family organisation seems to show important constraints on the organisation of the firm itself. As the years go by, a central role for the family is still reserved. The family plays a major role in this industry, not only in the early years but also throughout the life of the firm. There is evidence that the biggest changes in the business come from family members as can be seen in the following statement: Until 1974 the firm had 3 or 4 employees, as was common at that time. Meanwhile, my older brother started his university degree in business management and started helping my father. My brother mostly ran the business until 1994, the year I started working in the firm. . . . With my entrance into the firm, we started an industrial development process and built up a new factory. By that time we had 20 employees and about 550 m2 and now we have about 110 workers in a 5000 m2 area. (AM: 11)
The above information is important not only in terms of understanding the role of the family in the firm’s growth, but mainly to show generational management differences, an issue to which we will return later. The events organisation industry Within the events organisation firms, the process of entering the industry is quite different from the mechanisms described above. The events organisation industry is classified as a ‘recreational, cultural and sporting activity’ (NACE 92). In most cases, the entrepreneur’s background is in a related yet different industry. In many cases the advertising–marketing services industry was the source of potential entrepreneurs to enter the industry, either via advertising–marketing firms or marketing departments in a firm in a different industry. This is echoed by the following quote: I have always been connected to the marketing/advertising industry, where I was working in the sales department, and then I worked in an advertising firm, in which I had to organise a lot of events. (CD: 71)
Another different entry strategy into the industry is for the entrepreneur to start up the activity informally in the course of his or her personal life and then to upgrade it to a professional basis, as the entrepreneur sees a market opportunity, as described in the following quote: I started it as a hobby, organising the birthday party of the child of a couple who are my friends and since then things have been moving on . . . information
96
SMEs in a globalised world gets passed on from one to the other and eventually you get phone numbers of special paints suppliers, actors and balloons suppliers, etc . . . so the events area started on the basis of fun. (CD: 4) I used to work for a company and all my life I organised holiday programmes for friends and, at some point, I decided to materialise the idea and to create my own business. (SL: 4)
Both quotes illustrate the informality behind the idea of starting-up a business activity based in everyday life. The importance of weak ties, through the social network of the entrepreneur, in promoting an interest for entering the sector is shown, although in an indirect way. As shown in the first quote, friends become central to the idea of entering the industry, although their behaviour was not aiming to push the entrepreneur to enter the market (thus it was an indirect effect). In line with the theory of weak ties (Granovetter, 1973), the contacts from previous employment are a valuable asset in terms of providing new and important information, assuming these ties are mostly weak. The last quote clearly illustrates the importance of these ties, and how previous work experience can be a good source of new information and opportunities. Moreover, event organisers have previous work experience in different but related sectors. Coming from a different sector provides these entrepreneurs with more innovative (diverse) information. Entrepreneurs who come from the same sector are endowed with relevant information; however, this information is often redundant. Summary: similarities and differences across the two industries With regard to the process of entering markets, the two industries, show few similarities, although there are important differences to be pointed out. In both cases, most of the firms started-up on an informal basis, and then they became upgraded to a professional level. The difference comes in the type of job the founder does after work. In the case of furniture entrepreneurs, they replicate what they usually do in their daily ‘formal job’, while events organisers develop an expertise into different areas, by venturing into different and connected activities. The importance of informality and social networks in Portuguese new small firms appears very clearly from the examination of the two case studies. Nonetheless, this is part of an evolving mechanism in which the role of the family is decreasing as the society assumes a more urbanised attitude. The biggest difference between the two industries comes when referring to family involvement in business. While in the furniture industry, the family plays a major role in every moment of the firm, evidence shows
Family firms in the Portuguese furniture and events industries
97
that events organisation entrepreneurs rely more on weak ties, and are thus able to obtain more innovative information when compared to the furniture producers. These weak ties are sometimes responsible for the idea behind the creation of the firm, or at least they provide information to back the process of entering the market. Hence, at this point, the argument is that differences in both sectors can be partially explained by the balance of both types of ties. From a different perspective, the social and economic environment may also explain some of the differences between the two industries. Events organisation entrepreneurs face a more competitive labour market in terms of recruitment because they are mostly located in large cities, with competitive markets. On the other hand, there are significant differences in terms of local culture (urban versus rural environments), and with regard to the time period in which the firms entered the market. A large number of the furniture firms entered the industry in the period right after the revolution (in the 1970s), when despite economic uncertainty, expectations in terms of economic growth were high. The word ‘unemployment’, as a way to describe business start-ups, was never mentioned in the interviews of furniture firms. In fact, until recently, the northern coastal area of Portugal (particularly the region and adjacent regions of the furniture main production centres) has been characterised by unemployment levels close to the natural rate of unemployment, in the sense of Phelps (the stationary rate, or rate at which there is neither deflation nor inflation (Phelps, 1967)). Hence, the demand for labour in furniture production was abundant, whereas there was low or inexistent competition for work positions. Hence the decision for a business start-up was taken on the basis of entrepreneurial attitudes and on the expectations of further income and wealth (when compared to an employee situation). The fact that the events organisation industry is of relatively recent vintage in Portugal does not, unfortunately, provide the opportunity to evaluate and compare the basis on which businesses have subsequently grown. Financing An important issue in a firm’s survival is funding in the early stage of its life cycle. While it is a key issue in all stages of the firm’s development, it becomes a critical issue at the moment of the firm’s foundation. For the purpose of this research, the topic of funding is seen as providing a source of interesting insights since it is expected that entrepreneurs in the furniture industry rely more on family ties when it comes to this matter. The financing of activities in developed furniture firms cannot directly be compared with the funding of events organising firms, given that, at the
98
SMEs in a globalised world
time of this research, the two industries were at different levels of growth and maturity. Another reason why this point is highlighted right after the issue of business start-up is because the data collection focused on funding in the start-up period, rather than on how funding mechanisms operate in later stages of the firm’s activity, as the start-up period is seen as particularly important. The aim of this subsection is to analyse the extent to which funding problems trigger (conversely limit) cooperative relationships; this will help us understand how the family contributes to funding. Concerning the furniture industry, two main forms of funding were identified: family funding and the entrepreneur’s own savings. The main source of funding is family, most frequently the entrepreneur’s parents. The importance of these strong ties goes beyond the provision of cash as other types of resources were found to be provided as well. Yes, there was a time that my father borrowed money from my grandfather who was an emigrant in Germany. It was on the 25th of April5, . . . and we are talking about the German currency against the Portuguese one, which, by that time, had almost no value in international markets and that continued to depreciate steadily in the following years. (PT: 32–74)
As explained previously, working for another firm constituted a push to start-up a business in the furniture industry. Previous employment not only provided contacts, know-how and knowledge but also the initial capital to set up the business. This was particularly important in the mid-1970s (after the revolution year of 1974), when there were high levels of economic uncertainty; growing inflation rates and high interest rates made banks a very expensive and hence unattractive option for funding. Funding the firm through personal savings (from previous jobs and from the informal activity of production) was therefore one of the few options to start up a firm. Funding was always done with our own money. That’s why the firm grew very slowly. (PJ: 40) Today, my father reminds us of those times when he needed money to take us to the doctor and he didn’t have it, because all the money was invested in the firm. (PP: 42)
The second statement also shows that there is no distinction between money for personal use and for business purposes. This demonstrates how closely related family and business activities are, in a family-based firm. This may result in limitations to the firm growth due to funding limitations
Family firms in the Portuguese furniture and events industries
99
brought about by family involvement. Moreover, the financial strategy of the firm needs to be analysed considering the family as an extra activity which also needs to be funded by the firm. Moreover, a rather different source of funding was identified as being very important for the development of firms at a later stage of development. The integration of Portugal in the European Union provided a large number of funding opportunities for entrepreneurs to develop and upgrade their businesses. The aim of inquiring about this issue was primarily to understand the role of EU cohesion funds for the purpose of financing small firms in the European periphery, but it also provided interesting insights in terms of funding. In fact, for established and mature industries (as in the case of furniture), business associations played an important role in providing information about European support programmes. A number of firms have actually taken advantage of these opportunities by sourcing much needed funding that allowed them to expand their business. We applied three different times to IAPMEI6 supporting programmes. (AM: 77) We applied for all the European support frameworks, when there was support for the manufacturing sectors. Portugal joined the EEC in 1986 and in the 1990s there was a boom in terms of European support and my father applied for a bunch of them. (PT: 82)
The data suggests that firms with key linkages to business associations were more likely to effectively take advantage of EU funding, as tight relationships are key elements for access to information. Furthermore, firms that referred more strongly to this issue of funding are also those that have a role in the business association or have a strong connection to the local government, that is, the firms that are more embedded within the local productive tissue. The following quotes demonstrate the role of tight relations to business associations in accessing key people within these associations. The person who helped my father with the funding applications is someone who worked for some years for AEP7 and that was working in the International Relations and Business Support departments; thus, he had a great knowledge of the European framework programme and the business supporting tools. He was born locally and at some point he came to work in the local business association. Because my father was always involved in the business association since its creation, he also had a close relation to this person, who despite working for the business association, also owned a private firm for European funding applications. He had an advantage because in the beginning no one understood how it worked . . .. And there was also a friendship. (PT: 103)
100
SMEs in a globalised world
This is also important as it represents one of the few references to weak ties in the furniture industry interviews. The sector of events organisation definitely relies much less on family funding as well as on strong ties to obtain financial support. In fact, during the process of interviewing, it became quite clear that the family plays a limited role in the business, since references to family were never made on a spontaneous basis, except in one case when interviewed about funding the business: We [the four business partners individually] had to rely on family loans to be able to pay our suppliers, but it was a very short-term situation, and after three or four months the money was paid back. (3L: 139)
In most cases, previous employment provided the necessary financial resources for the business, as described in the following quote: We started out with our own capital, mainly because at that time, there would be no one who would provide us with the necessary financial resources for a business like this one. (IN: 136)
When it comes to finance in a firm’s start-up, various industries provide different mechanisms. The first main difference is of a structural nature and it relates to different levels of entry barriers into the respective markets. While furniture firms need to rely on heavy investment in terms of machinery, events organisation firms can start-up with a very small level of funds. Moreover, furniture firms rely on the family for the start-up of the firm and on government supported programmes for the firms’ development, but events organisers rely on these government supported programmes for the early stage of the firm. However, this fact needs to be contextualised in the different periods in which both industries emerged. In both cases, previous employment is a source of funds, for the firm’s start-up. Figure 5.2 shows how both industries access funding. A partial explanation for the fact that both types of firms access supporting programmes at different stages in the firm’s development is that these programmes became popular in the 1990s when furniture firms had already gone through the start-up phase. As previously explained, furniture entrepreneurs tend to rely on strong ties in the early years of the firm that, to some extent, limited their access to information about these programmes. The combination of these two facts explains the reason why supporting programmes only become relevant in a more mature stage of the business. Moreover, the existence of local business associations provides an opportunity to easily access this funding, if information is obtained via them. The data explains how these ties (people in business associations)
Family firms in the Portuguese furniture and events industries
101
Furniture 5.00
Events organisation
4.00
Value
3.00
2.00
1.00
0.00 Own funding
Family
Banks
EU/other government programmes
Notes: The value for EU/other government programmes for the furniture industry does not represent funding for the firm start-up, otherwise the value would be 0. The values represent the number of firms accessing each type of funding source. The value represents the number of occurrences in the interviews. Source:
Elaborated by the authors.
Figure 5.2
Funding sources
were important in accessing funding information. Concerning the events organisation industry, only general programmes have been made available to date. As evaluation and management of these programmes is made centrally, in government head offices (instead of dispersed over the country), they are less well adjusted to specific regional circumstances and thus aim to support firms in a general manner. Finally, another key element in understanding furniture firms’ financing strategies is the role of the family. There is evidence suggesting that the boundaries between firm and family are very weak and blurred, leading to situations that could not be fully explained by the assumption of economic rationality. The furniture firm financing strategy embodies family financial needs, which are not solely based on business considerations. This rather
102
SMEs in a globalised world
limits the growth of the firm as external investors, who could potentially bring additional funding to the firm, face high barriers when trying to enter the firm. Hence, it can be inferred that furniture family-based firms limit access to opportunities for cheaper funding. Interestingly, funding through the banking system was only referred to by one of the firms, which shows that bank funding is rarely a source of funding for both industries (and this maybe generalised to other industries in Portugal). Several reasons can be pointed out, that explain this état-defait: 1. 2. 3. 4. 5.
the traditional inefficiency of the Portuguese banking system to provide adequate funding solutions;8 bank funding strategies are limited to firms after a growth period, and do not concern firms in the early years; the lack of trust in the banking sector; since relying on bank finance might not be important from a cooperation perspective, it was not referred to in the interviews; and/or the appearance of EU funding makes bank funding a second-best option.
The emphasis of this subsection was mainly on financing in new small firms. Perhaps, different conclusions would have been reached if the collected data had concerned firms’ funding at a later stage of their development and growth strategy. However, what should be stressed is that cooperation and external contacts to the firm are present and are important for the issue of funding business start-ups. Advantages can be found in contacts with family and in weak ties (to access information about funding). It becomes evident again that the furniture entrepreneurs rely more on strong ties, and that events organisers rely on weak ties. Although the entrepreneurs’ savings are a source of funding they do not represent a cooperative mechanism.
OTHER KEY ASPECTS OF FAMILY-BASED FIRMS Getting the family involved in a firm brings very particular properties to the nature and characteristics of business, which provide both gains and limitations to firms. The first part of this section aims at understanding the impact of family-based relationships on network and collaborative behaviour in business development, and the extent to which these relations evolve as the firm becomes more mature. The second subsection deals with the case of ownership succession.
Family firms in the Portuguese furniture and events industries
103
Evolution of Family-based Relationships Table 5.1 summarises some of the relevant advantages and limitations of involving the family in the firm, and it functions as a guide to the rest of this section, where these issues are analysed in greater detail. The literature suggests that strong ties are more important in the first years of the firm’s existence, and as time goes on, these ties give way to weak ties (when more new information is brought in). The data collected from the furniture industry is rich in providing examples of the importance of the family during the first few years of a firm. In most cases, the family contributed with funding in the early years of the firm, mostly with personal savings from the entrepreneur’s parents. The results strongly suggest that the furniture industry relies more on family relations than the events organisation industry. The lack of reference to the family in the case of events organisation (even when inquired about its role in the business) shows that the family plays only a marginal role. Hence, in order to describe the role of the family in cooperative relationships, one has to rely on evidence collected from the furniture industry. However, due to similar features permeating the traditional industries, this section can thus be used to describe the role of the family in Portuguese Table 5.1
Advantages and limits of a family-based firm
Advantages
Limitations
Provides access to a wider net of social contacts, through the different family members Provides resources to the firm (money and other physical resources) Gives access to a pool of labour
Limits the access to external investors thus limiting access to a cheaper way of funding the firm Defines a hierarchy non-based on professional competences Brings familial problems into the core of business activities Is less exposed to innovative information
Access to information about potential employees Has a reliable and motivated workforce The increased trust level permits higher interaction and it facilitates cooperation, giving access to an easier division of labour that increases the firm competitiveness Source:
Elaborated by the authors.
104
SMEs in a globalised world
traditional industries and to demonstrate that the modern industries rely on other actors than the family. In family-based firms, the social network of each member of the family is important, and hence a family business is more likely to have a wider network of social contacts, that may represent a bigger pool of customers.9 By contrast, it can also be suggested that strong family links may hinder the establishment of new contacts outside the family, thus rendering the social network of each family member smaller, when compared to actors relying less on family ties. The fact that furniture entrepreneurs are more dependent on the family and that the family plays a more important role in the business, may be due to differences in terms of cultural and religious values. The great interaction between family and business is found in most traditional industries, which are located in rural and semi-rural areas. The presence of the family in the business seems to be a consequence of social mechanisms, although these are becoming less evident. If one takes the events organisation industry as an indicative trend towards the emergence of modern firms in Portugal, the results demonstrate that this is a part of a more generalised trend of decreasing importance for the involvement of the family within the business. Another important issue, in terms of the family role in the business, is the family reputation, which is very apparent in small geographical areas. In these areas, it is common that a family is known through its members’ attitudes towards life, honesty and other values. Reputation also plays an important role in the recruitment of new employees. The family of the employee (through its local reputation) is a great indication of employees’ attitude. It is common to hear comments about the potential quality of an employee as he or she comes from a hardworking family. The gains arising from this phenomenon are related to decreasing the risk generated by the experimental recruitment of unknown employees (whose attitudes may not fit the expectations of the employer in terms of the creation of a firm identity). It is also common for the firm’s workforce to come from the family itself. Particularly in the furniture industry, a lot of evidence was found of several members of staff being related to the owner of the firm. Again, the same trust mechanism is in operation and the gains are as described above. The provision of factors of production to the firm by the family is also related to the physical fixed assets that can be inherited by the family firm. However, due to specific technological characteristics such as innovation, machinery becomes rapidly obsolete within a generation, and thus there is limited use for production equipment inherited from previous generations. Another main finding is that the family plays a major role in terms
Family firms in the Portuguese furniture and events industries
105
of the division of labour. Firms owned by family-related entrepreneurs provide relationships with higher levels of trust and increased cooperation. Cooperation is then eased between these firms as a family bond is at the core of the relationship, and the literature suggests that family values seem to predominate over business values (Howorth and Ali, 2001). Examples of firms owned by the same family members are rich in explaining the gains from this close interaction. The division of labour increases competitiveness, by allowing a narrower specialisation in the product or in the value chain, and by permitting smaller organisations (with lower organisation costs) to benefit from lower transaction costs. According to transaction costs economics, smaller and more specialised firms derive gains in terms of coordination costs (as larger firms enjoy decreasing returns to scale) but they also face higher transaction costs (due to the more frequent outsourcing of contracts). However, if outsourcing is done to family-related firms, these contract costs are importantly diminished as trust levels increase. Hence, the result is that family-based SMEs have both lower organisation costs and potentially lower transaction costs, when compared with large firms.10 However, the family also presents some limitations to the firm’s growth. This is due to the firm’s organisation and hierarchy based on seniority in the family, rather than on a definition of competence or management ability. In the same vein, family divergences may be brought into the firm, affecting business and shaping events that are not related to the business. The data also describes the blurred boundaries of funds allocation. As the finance of the family and the firm get confused, problems arising in one area affect the other. Moreover, such involvement of the family in the finance of the firm acts as a barrier to external participation in the firm and thus in the firm’s capital. One of the most practical ways of getting much needed funding is by using external shareholders. As the family acts as a barrier to the entry of external people into the firm, it also limits access to a cheap option of funding, but at the same time it guarantees that decisionmaking remains at the heart of the family group. The family is the principal source of values (both familiar and social) as a certain convergence of attitudes towards life is expected among family members. As the family lies at the core of the transmission of both business and social values, it is expected that business attitudes are also shared across the different generations, although adapted to changes in society. In that sense, a set of rules is shared by the various members of the firm (family) and to some extent, by the different generations, as family values seem to prevail over strictly business routines. Hence, the cohesion of the family determines the homogeneity of cultural values inside the firm.
106
SMEs in a globalised world
The Generational Conflict One way in which a rupture can be exercised in the twin tenets of family and firm is through the conflict opposing different generations at the moment of the succession of ownership. While in non-family-based firms, succession of ownership is determined by strictly legal and/or economic factors (such as capital ownership), in family firms ownership tends to be kept within the different generations. A potential conflict can impact upon the growth, and indeed the survival, of the firm. This subsection compares both first and second generations of entrepreneurs within the furniture industry and the events organisation industry in order to identify the changes in the entrepreneurial attitude and to evaluate how these changes impact on the business in general. Making a clear distinction also provides some insights in terms of the influence of urban and rural contexts in the way businesses are conducted as the type of entrepreneurs are categorised accordingly. For the purpose of this analysis, one should consider first generation furniture entrepreneurs as settled in a rural context, whilst events organisers operate in an urban context, and second generation furniture entrepreneurs are semi-rural entrepreneurs. This analysis can also be seen as an evolutionary perspective of the entrepreneurial attitude in Portugal. The fact is that the recent urbanisation of Portuguese society in general, stirred by EU integration, has changed the business society accordingly. Although first generation furniture entrepreneurs have witnessed structural sociological changes, rural institutional values remain as the main source of entrepreneurial action, at the core of their behaviour. This means that although not behaving in the same way as three of four decades ago, some of the embedded cultural values (such as trust or honesty) remain unchanged. For example, a second generation entrepreneur explains the extent to which first generation entrepreneurs display a low appreciation of business associations, since ‘in their time they could survive without business associations’ (LB: 15). Second generation entrepreneurs in the furniture industry are considered as operating in a hybrid type of urban– rural environment for two reasons: 1.
2.
these second generation entrepreneurs, although brought-up and educated with rural values have, generally spent time in urban areas getting their university degrees; and the second generation is composed of individuals who were born at a time of easier access to urban areas (due to the recent development of transport infrastructure); they are more exposed to globalisation brought about by increased communication such as the Internet
Family firms in the Portuguese furniture and events industries
107
which makes their contact with the external world easier, and where information is diffused more rapidly and easily. In the furniture industry, and in line with the findings derived from the previous section, business development seems to be highly constrained by family evolution. Thus, the firm’s development usually depends on the second generation’s ability to take over the management of the firm and to introduce new management skills. During the course of the interviews, the differences between the two generations were sought, as a way to understand in what way the entrepreneurs’ age and education affect the way they see business. During the process of data collection, important similarities were also found between second generation furniture and the events organisation entrepreneurs. This fact is very important in supporting the argument that age and educational levels are important factors in the explanation of the different approaches to cooperation. Within the furniture industry, the differences between the two generations are as follows: 1. 2.
3.
4.
the second generation is usually more innovation and design driven and more open to R&D; although there seems to be evidence of informal cooperation in the first generation, the second generation is more inclined towards the formalisation of cooperation and is more willing to enter cooperation support programmes; a bigger emphasis is placed on market and firm image by the second generation management, while the first generation is more concerned with production aspects; the second generation is more open to business associations taking an active role in business and is more willing to use the services they provide.
Including the second generation of entrepreneurs in this analysis is necessary as these entrepreneurs greatly influence the decisions in the firm and some of them are even at the top management level of the firm, despite the differences referred to previously. Moreover, this analysis permits a more accurate view of the potential cooperation characteristics between the furniture firms. Possible conflicts may arise when management changes from one generation to another and both are working in the firm. When this happens, it is always the second generation that is constrained by the first one. The first generation is in a powerful position, which reflects power asymmetries inside the family. In any case, the second generation recognises that it still
108
SMEs in a globalised world
needs to be backed-up by the knowledge of the first generation (regarding production and the ‘tricks’ of the business), as well as the network of contacts built-up over time. It seems clear that despite the generational conflict in family firms (often quoted in the literature), there are also advantages arising from the coexistence of both generations in the firm’s management as can be seen in this quote from a second generation furniture entrepreneur: In the firms in the succession stage, the second generation entrepreneurs want to change, but in a way they are limited by the conflict between the generations. New firms, created by entrepreneurs of my generation (lacking the support of the first generation) gamble heavily on more modern management, but most of the times it goes all wrong, the market is very tricky . . . in terms of competition one needs to be aware of all the tricks . . . that come from the experience of the first generation . . . but it is also in firms that have experienced succession [compared to those still relying only on first generation management] where modernisation has happened and there is a higher importance on the firm image, marketing, quality of service, and so on. (LB: 325)
It seems that in the case of family-based firms, the firm becomes a repository of knowledge and information and hence it is more adapted to market changes, along with enjoying the advantages of first generation experience. However, it must be noted that conflicting positions also act against firms with successive generations. All in all, second generation furniture entrepreneurs are in a better position than the event organisers and take advantage of both formal and informal collaboration as the pool of contacts seems to be wider.
CONCLUSIONS This chapter has explored particular types of SMEs in the periphery of the EU: that of family firms in Portugal, a country with core traditional values in society that has nevertheless undergone a substantial number of transformations in the past 30 years or so. The chapter deals with the issues of firm creation, financing and succession across different generations of family firms in the context of past and new challenges brought about by integration and globalisation. It has used a comparative analysis to explore the contribution of family business to the evolving social and economic transformation of Portugal. The furniture and events organisation industries were used as an indication of the evolution of business in Portugal. In particular, the first generation furniture industry, an industry that displays a traditional approach to business management, and the
Family firms in the Portuguese furniture and events industries
109
events organisation industry, which is based in a more urban and modern perspective, were the main focus of the study. Second generation furniture entrepreneurs, lying in-between the two, were also analysed. Since the emphasis was placed on family-based firms in the furniture industry, the events organisation firms were primarily used with a comparative purpose in mind, and were thus marginalised, to some extent, in terms of reporting the results. A series of semi-structured in-depth interviews with 29 organisations (including some local and national supporting institutions) were conducted in 2005 and 2006. Most of the firms interviewed are small firms and were founded after the Portuguese revolution of 1974. They therefore underwent a number of economic transformations characterising the Portuguese economy; those transformations culminated with both EU membership in 1986 and globalisation. With the dual aim of examining the nature of business when the family is at its core and the mechanisms of family management succession, this chapter brings a number of salient conclusions. After clarifying some definitional issues and after discussing briefly the advantages and disadvantages of organising production within the ambit of the family, as seen in the literature, it was shown that, in line with the institutional economics perspective, the analysis of family firms needs to be contextualised within a cultural framework that determines the social nature of the family and that influences therefore the way family business is conducted. In this respect, core traditional values of Portuguese society (such as the importance of the religion, hierarchy in the family and so on) seem to permeate daily business life more in traditional industries such as the furniture industry, and particularly in the past, than in modern industries such as the event organisation industry. In particular, the family has been playing a much bigger role in the case of furniture firms at the start-up stage with regard to financing; it continues to play such a role in terms of both recruitment of employees and also of involvement in socio-economic networks. The role of the family is however less strong for second generation furniture firms. By contrast to the first generation furniture firms, events organisation firms rely more on weak ties; in particular, they face a more competitive labour market for the recruitment of their employees, a factor which is also due to the fact that they are more urbanised when compared with furniture firms. This distinction between strong financial ties for furniture firms and weak ties in the case of events organisation firms also appears at the level of financing in the early stages of the firm. One common feature among the two industries is however that they only rarely rely on banks’ financing. The last section discussed the significance of strong versus weak ties in the context of the firms’ growth and evolution, and it analysed the
110
SMEs in a globalised world
problem of ownership succession. A generational divide appears clearly in furniture firms, with second generation entrepreneurs behaving in line with events organisation entrepreneurs. When conflict arises between two generations at the moment of management change in furniture firms, it appears that seniority prevails and that the second generation gives way to the first generation. Consequently, this study hints towards the evolution of Portuguese SMEs, be they from traditional or less traditional industries such as the events organisation industry, into ‘modern’ business entities with a number of characteristics. First of all, the family in these SMEs assumes a decreasing participation in the business, a decreasing role compensated for by the development of the SMEs’ network extending to different actors, such as business associations and national or supranational entities for the purpose of cooperation in the case of support programmes. Second, these new types of Portuguese SMEs place a greater emphasis on marketing issues rather than on purely production issues, and they are more innovation and design driven than the SMEs of the past. It seems, therefore, that the once convenient distinction between traditional and modern industries is blurred, and that SMEs in Portugal are catching up with their EU counterparts. Being more aware of external influences, these SMEs have increasingly been responding to the forces of economic integration and of globalisation. Although not commonly studied in the literature, a different perspective would be to explore interfirm relationships between businesses conducted by different family members. Although the firms may be independent institutions per se, there is a strong familial link between the entrepreneurs. In line with the theories presented in this chapter, it seems clear that these relationships are different in their nature from those established by firms with no family connection.
NOTES 1. 2. 3. 4. 5.
For a brief account of the literature on family firms, the reader can refer to Amann and Jaussaud (2008). For an interesting discussion on the definitional issue, and the way different definitions of family firms might affect the findings of the different studies, see Westhead and Cowling (1998). Located in the North of Portugal, and known as one of the centres of national furniture production. ‘Barraco’ can be translated as shed, but the word embodies different meanings. It is in a way associated with the idea of a shabby and dirty detached house, garage or stable. The 25th of April represents the day of the Portuguese revolution in which major changes occurred: political, social and economic. It represents the end of the dictatorial
Family firms in the Portuguese furniture and events industries
6. 7. 8.
9.
10.
111
regime and the end of the colonial era. The years following the revolution were characterised by high levels of uncertainty. IAPMEI, Instituto de Apoio ás Pequenas e Médizs Empresas e ao Investimento [Institute for the Development of SMEs and for Investment] is a Portuguese SMEs supporting institution. AEP: Associação Empresarial de Portugal (Portugal Business Association). There is a clear reason for not accessing bank loans in the 1970s and 1980s as interest rates were too high for investors. However, the decrease in interest rates was theoretically expected to result in an increase in bank funding, a phenomenon that did not take place, suggesting an inefficient national banking system. In the events organisation industry, the data contains the example of a firm who mostly works for the pharmaceutical industry. There is a family link to the pharmaceutical industry, and hence, access to the customer is gained via the family. The family also provided the entrepreneur with premises for the business, in the same way that the social network of the entrepreneur contributes to the customer base of the business; in a family-based firm, the social network of each family member becomes a target for business. Lower costs in small family firms are illustrated by the following case where five sons and a daughter of a former furniture entrepreneur created five different firms in the furniture industry. Although competing in some markets, one of the entrepreneurs (the daughter) admits that between all five firms there is no need for written contracts for business-related transactions. From time to time, each pair of firms (and not on a regular basis) accounts for the amount one has to pay to the other. In that sense, costs associated with completed contracts are zero; hence it becomes less costly to establish transactions between themselves than with firms outside the ‘family network’. This case is a clear example of lower transaction costs between family-related firms.
REFERENCES Amann, B. and J. Jaussaud (2008), ‘Family business in Japan: an overview’, in B. Andreosso-O’Callaghan and B. Zolin (eds), Asia and Europe: Connections and Contrasts, Venice: Ca’ Foscarina, pp. 187–202. Chrisman, J.J., J.H. Chua and P. Sharma (2003), ‘Current trends and future directions in family business management studies: towards a theory of the family firm’, Coleman Foundation white paper series, Chicago. Donnelly, R.G. (1964) ‘The family business’, Harvard Business Review, 42 (4), 93–105. Gallo, M.A, H. Jenkins, I. Máñez and K. Cappuyns (2002), ‘Internacionalizacion via Allianzas Estrategicas en la Empresa Familiar’, documentos de investigacion no. 477, Barcelona: IESE. Gattrel, J., H. Jenkins and J. Tucker (2001), ‘Family values in family business’, in G. Gorbetta and D. Montermelo (eds), The Role of Family in Family Business, 12th Annual FBN World Conference Research Proceedings, Milan: EGEA. Granovetter, Mark (1973), ‘The strength of weak ties’, American Journal of Sociology, 78 (6), 360–80. Howorth, C. and Z. Ali (2001), ‘Family business succession in Portugal: an examination of case studies in the furniture industry’, Family Business Review, 14 (3), 231–44. Phelps, E. (1967), ‘Phillips curves, expectations of inflation and optimal unemployment over time’, Economica, 34 (3) (August), 254–81.
112
SMEs in a globalised world
Sharma, P., J.J Chrisman and J.H. Chua (1997), ‘A strategic management of the family business: past research and future challenges’, Family Business Review, 10 (1), 1–35. Venter, E., C. Boshoff and G. Maas (2005), ‘The influence of successor-related factors on the succession process in small- and medium-sized family business’, Family Business Review, 18 (4), 283–303. Watts, G. and Tucker, J. (2001), ‘Learning the family way: early learning experiences within the multi-generational firm’, proceedings of the 1st International Entrepreneurship Forum in Entrepreneurship and Learning, June, Naples, Italy. Westhead, P. and M. Cowling (1998), ‘Family firm research: the need for a methodological rethink’, Entrepreneurship Theory and Practice, 23 (Fall), 31–56. Westhead, P. and D. Storey (1997), ‘Training provision and development of small and medium-sized enterprises’, Department for Education and Employment research report no. 26, London: HMSO. Zahra, S.A., J.C. Hayton and C. Salvato (2004), ‘A resource-based analysis of the effect of organisational culture’, Entrepreneurship Theory and Practice, 28 (4), 363–81.
APPENDIX 5.1 Table 5A.1 Firm Code LB AM PP PJ CD SL PT 3L IN
Characterisation of the firms interviewed Furniture
Events organisation
X X X X
Familybased firms
First generation
X X X X X X
X
X X X
X X
X X
6.
Forms of industrial development in Chinese specialized towns and types of challenges to European manufacturing SMEs: an Italian perspective Marco Bellandi and Annalisa Caloffi1
INTRODUCTION The rise of ‘made in China’ has been favoured by international traders and multinational firms around the world. It has also aroused the concern of many SME (small-to-medium sized enterprises) manufacturing clusters in developed countries, for example in Italian industrial districts and in less developed countries as well. The Chinese competitive challenge comes from a rich variety of factors. They include of course cheap labour and land; a large internal market; state and local policies opening to foreign direct investments (FDIs) by multinational firms with increasing capacities in the management of international production systems (Gereffi and Korzeniewicz, 1994; Arndt and Kierzkoswki, 2001). Also, recent investigations have pointed out the role of local reserves of entrepreneurship and competence, adequate regional policies supporting the development of industrial clusters and the constitution of a large educational infrastructure, research and communication, as well as the influence of networks of overseas Chinese entrepreneurs and Asian international value chains (Enright et al., 2005). A laboratory for understanding this variety is given by several cases of specialized towns which have been developing in the hotbeds of Chinese new industrialization, such as in the southern province of Guangdong. The second section opens up with a summary of previous work on this topic, and it relates this work to studies and models of local development built around the experience of Italian industrial districts (IDs).2 The issue of industrial development forms in Chinese specialized towns and the way this can be informed by the experience of Italian IDs deserves attention 113
114
SMEs in a globalised world
for at least three reasons. First, the reference to models of local development helps in generating hypotheses on variables which may be relevant for assessing the nature and variety of forms of industrial development in the Chinese specialized towns. Second, the application to the peculiar experience of Chinese new industries, even when possible, reveals interesting variations of those models. Third, the extension and fragmentation of production systems and value chains managed by multinational firms is a general tendency which is hitting the SMEs industrial clusters of many countries (Humphrey and Schmitz, 2002) as well as of Italian industrial districts. Their SMEs and the related collective/public–local/regional agencies are learning how to react progressively to this challenge, and the reference to the Chinese experience helps to understand better, both the nature of the challenge and the strategies of competitive and collaborative reaction. Building on those premises and on an appropriate fieldwork, including a collection of official statistical data and the scanning of various sources of complementary information, a ‘cluster analysis’ has been carried out. The data refer to a set of 66 specialized towns in the Guangdong province. The hypotheses, the definition of operational variables, the resulting typology, and some comments on the identified forms of industrial development are illustrated in the third and fourth sections. The fifth section reports details referring to a number of case studies. The data and the case studies refer to the years 2003–05: by which time the local conditions would have changed in quantitative and qualitative terms. We maintain however that the results illustrated here identify relatively general types of industrial configuration, which characterize phases of industrial development within the Chinese specialized towns, with related evolutionary potentialities and competitive capacities. The sixth section deals with the relation between the typology of specialized towns, the different challenges that they have raised to European SMEs industrial clusters, such as in the Italian industrial districts, and the strategies of competitive and collaborative reaction that may be enacted in such clusters and districts.
CHINESE SPECIALIZED TOWNS AND LOCAL DEVELOPMENT If we consider the Chinese challenge just as coming from a myriad of local and international firms producing mass manufacturing standardized products in China by exploiting a large supply of cheap labour and land, then the reaction to the Chinese challenge tends to be played at the level of the strategies of delocalization of single firms and at the level of national
An Italian perspective of Chinese specialized towns
115
governments (and the EU) negotiating protection against the worst forms of price competition and imitation.3 The approach and the design of reaction strategies change when it is understood that: (a) the more solid Chinese challenge to SME clusters, like those in the Italian industrial districts, comes from ‘new industries’; (b) those new industries are characterized by capacities of local entrepreneurship, wide strategies by international business and diffused public support combining various levels; and (c) these combinations operate diffusely at the level of several clusters of firms, such as in the so-called ‘specialized towns’.4 Specialized towns in Chinese industrial regions appear to be ‘localities characterized by a significant agglomeration of industrial activities and by a high share of them concentrated within localized industry’ (Bellandi and Di Tommaso, 2005, p. 708). A similar definition is explicitly assumed, for example, by the provincial government of Guangdong.5 The core of the province is the Pearl River Delta (an area stretching between Canton, Hong Kong and Macao), which is one of the most industrialized areas of China (Cheng, 1998; Enright et al., 2005). The definition is applied within provincial policies aimed at supporting innovation and quality improvement in manufacturing SMEs. Similar phenomena and definitions are found elsewhere in China, in particular in the other industrial powerhouse, that is, the area of the Yangtze Delta, around Shanghai.6 Do the clusters of the specialized towns correspond to something more than random and temporary agglomerations of enterprises? May we find endogenous characters, related to social, industrial and public support? Such questions are generated by reference to models of local development, such as those exemplified by many industrial districts in contemporary Italy (Becattini, 2004). In quite synthetic terms, it could be argued that local development is characterized by an evolving coupling between a ‘local production system’ and a ‘locality of industry’ (Bellandi and Di Tommaso, 2006). The first is intended as a cluster with systemic characters, where sets of specialized firms, usually SMEs, run complementary activities, possibly extending to various manufacturing and service sectors related to the core business field; the organization and development of the cluster is supported by the fabric of social and civic life of a locality of industry. This is a locality with a socio-economic identity (and corresponding to a set of contiguous towns, villages, rural areas, with a principal town or city) characterized by the presence of one or a few local production systems. Outside ideal–typical conditions, local development is more generic, less endogenous, more centralized. Beyond certain thresholds of sectoral dispersion among the agglomerated activities, of dependency from external strategic control,7 and of business centralization, clustering in a place is
116
SMEs in a globalised world
just the effect of the variable location choices of more or less footloose companies.8 In Guangdong, the role of foreign multinational firms is extensive, and the presence of clusters supported by local forces of development is less obvious, especially in the areas of the Pearl River Delta near Hong Kong, like in the prefecture of Dongguan. However, some cases, more characterized by local forces, have been assessed in other areas of the Delta region, such as in the Foshan prefecture (Qiu and Xu, 2004; Long, 2005). Here, local business traditions, supporting the industrial take-off, are found in some specialized towns, especially where textile, clothing, shoes, furniture, ceramic industries and the like predominate. Electronic clusters in specialized towns, and generally the industrial clusters located within larger cities, are more often based on external investment by large firms, looking not only for cheap labour, but also for easy access to good transport systems and logistics as well as trade channels. In Zhejiang, another booming Chinese province, and in particular in the Wenzhou prefecture, the strength of clusters producing for internal markets has been related explicitly to the role and presence of widespread traditions of local entrepreneurship, rural craft and trade skills, SMEs, all these showing important similarities with models of Italian industrial districts led by SMEs industrial clusters (Christerson and Lever Tracy 1997; Wang and Tong 2005). In a third booming province, Jiangsu, and in particular in the Sunan region, near the autonomous municipality of Shanghai,9 clustering has been characterized more by traditions of collective manufacturing enterprises (Biggeri et al., 1999).10
A MULTIVARIATE ANALYSIS ON SPECIALIZED TOWNS IN GUANGDONG This section introduces a statistical analysis of a data set on specialized towns in Guangdong, in particular in the Pearl River Delta. The analysis points to the identification of various forms of industrial development at the local level in this area, taking the model of local development led by SMEs clusters, as in Italian industrial districts, both as the basis for the elaboration of pertinent variables, and as a guide in the interpretation of the results. A set of 72 specialized towns has been defined for investigation within the ‘China and Italy Research and Learning Project’.11 They include those officially acknowledged in 2003 by the Guangdong Provincial Government and other cases whose identity and importance has been cross-referenced. Out of this set, 66 cases have been selected for our statistical analysis, considering the availability of reliable data.
An Italian perspective of Chinese specialized towns
117
The set of 66 specialized towns observed here encompasses a total population of around 2.6 million inhabitants (that is around 3 per cent of the total population of Guangdong Province, with around 40 per cent of them being migrant workers), and a total of 64 257 enterprises (China Township and Village Statistics data).12 Data collected from various sources,13 and referring to the years 2003 to 2005, are used for defining indices which are elaborated through a ‘cluster analysis’. The indices approximate roughly a set of socio-economic features related to the model of local development recalled in the second section. The groups identified by the cluster analysis thus reflect the signs of different forms of industrial development among the specialized towns, more or less related to local development. The next sub-section shows the results of the analysis, while the following paragraphs and Table 6.1 illustrate the meaning of the data and the indices. Locality of Industry A first feature is the weight of the local population as a resource for the industry. The migrant population14 is here assumed as a proxy for migrant workers who usually are compelled to move from poor rural areas (usually from internal Chinese provinces) to find jobs in the richer coastal regions (Ma and Lin, 1993; Yao, 2001). A lower ratio of migrant population (MIGRPOP) suggests a higher involvement of the local population in the local specialized industry. This is a first sign related to the presence of conditions which in the second section have been associated with the idea of local development. Stronger localities of industry should present a fairly welldeveloped set of complementary activities around the manufacturing industry, in particular a set of business services. The proportion of the output value produced by the tertiary sector in the total value of the local economic activities is used here as a proxy. Given that the observed set of specialized towns does not include large urban areas (we are dealing mainly with small towns with a strong industrial character), a higher level of the index leads us to expect a larger development of business services (TERTIAR). Sectoral Specialization Each specialized town is coupled with a principal sector of activity, generally in a manufacturing industry, defined statistically at the second digit level. This sector is defined as the ‘specialized industry’, or as the industrial cluster characterizing the area. A relevant feature concerns the weight of the activities of the specialized industry within the locality. It is considered as the proportion of the industrial output produced by the
118
SMEs in a globalised world
Table 6.1
Indices and source of data
Indices
Description
MIGRPOP
The proportion of the migrant population to the total population of the town (year 2003). Source: China Township and Village Statistics (NBS, 2004a). The proportion of the output value of tertiary sector to the output value of the whole local economic activities (year 2003). Source: Di Tommaso and Rubini (2006) on Guangdong Department of Science and Technology (DST) data, and fieldwork investigation (see Note 13). The proportion of output value produced by the specialized industry to the industrial output value of the town (year 2003). Source: Di Tommaso and Rubini (2006) on DST data, and fieldwork investigation. The proportion of wholly foreign enterprises (WFO), JVs and other forms of foreign or sino-foreign enterprises to the total number of (non-micro) enterprises of the town (year 2003). Source: Di Tommaso and Rubini (2006) on DST data, and fieldwork investigation. Proportion of land area occupied by industrial parks, innovation park, and other economic zones on the total area of the town (km2) (year 2004). Source: fieldwork researches, local newspapers (China Daily, South China Morning Post, various years) and broader researches on the Internet (mainly local government websites). The proportion of specialized trade markets on total trade markets of the town (year 2003) Source: China Township and Village Statistics (NBS, 2004a) and fieldwork investigation. Number of innovation centres, research centres, technology transfer centres of the town (year 2004). Source: fieldwork researches, local newspapers (China Daily, South China Morning Post, various years) and broader researches on the Internet (mainly local government websites).
TERTIAR
SPECIND
FOREIGN
INDPARK
SPECTMK
INNOCEN
Source:
Authors’ elaboration.
specialized industry (in value) to the industrial output produced in the town (SPECIND). Weight of External Firms and Related Public Support Another industrial feature concerns the local role of external businesses. A first proxy is the proportion of foreign enterprises in the total number of
An Italian perspective of Chinese specialized towns
119
local (non-micro) enterprises (FOREIGN). Foreign enterprises predominate where their entities in the area are represented by large or very large plants. A complementary proxy is represented by the creation of industrial parks, usually targeted at attracting foreign investors or plants of big Chinese companies (INDPARK) (Eng, 1997; Chelan Li, 2001; Yeung, 2002; Walcott, 2003). Local Firms and Related Public Support A final feature touches upon the presence, extent, and nature of the fabric of local firms, usually SMEs, within the specialized industry. An aspect which is easily associated to a well-developed or developing fabric of local firms is the creation of innovation centres or other collective infrastructures devoted to the promotion of the innovation15 (INNOCEN). A second element is the access to external markets. One of the most common policies realized by local governments supports the strengthening or even the constitution of a specialized market, that is a local organization with appropriate infrastructure hosting trade facilities for industrial companies and traders, both local and non-local. They provide local SMEs with an alternative channel (with respect to the inclusion within the distribution networks of big clients or suppliers) to reach markets for purchasing inputs and selling products, sometimes only at a regional or national scale, sometimes at an international or even global scale (Wang, 2006). When the creation of this kind of organization is targeted towards a specific sectoral specialization of the local enterprises it takes the form of a cluster-based policy (SPECTMK).
FOUR TYPES OF SPECIALIZED TOWNS IN GUANGDONG Different combinations and different intensities of the above mentioned indices are the signs of different forms of industrial development among the specialized towns, more or less related to local development. The proxies have been elaborated through a multivariate analysis (of the ‘cluster’ type) aimed at defining a typology that is not purely descriptive, since the identification of the variables and the interpretation are led by the common reference to models of local development. The results are illustrated in this section. The statistical clusters are referred to as ‘groups’, in order to avoid any confusion with industrial clusters. The analysis leads to the identification of four ‘groups’ of specialized towns (Figure 6.1).16
120
SMEs in a globalised world
4.00 generic
3.50
inward
district
foreign
3.00 2.50 2.00 1.50 1.00 0.50 0.00 –0.50 –1.00 –1.50
MIGRPOP
Source:
Table 6.2
1 2 3 4 Source:
SPECIND
FOREIGN
SPECTMK
INDPARK INNOCEN
Authors’ elaboration.
Figure 6.1
Cluster
TERTIAR
Four groups of specialized towns in Guangdong (2003–05) (means of the seven indices in the four groups of specialized towns) Distance among the centres of the cluster Generic (1)
Inward (2)
District (3)
Foreign (4)
0
2.509 0
3.489 2.925 0
4.838 6.431 6.155 0
Authors’ elaboration.
Figure 6.1 displays the means of the seven indices in the four groups of specialized towns (standard values; k-means methods, see Note 16). The groups show different combinations of intensity of the previously identified indices, with a larger variance of the incidence of foreign enterprises and industrial parks, and of migrant population. On average, one group more markedly differs from the others (it exhibits the maximum distance from the others, see Table 6.2): it is the ‘foreign-driven’ group of specialized towns, strongly characterized by the presence of foreign enterprises and industrial parks. At the other extreme we find the ‘inward’ areas, which are characterized by the lowest presence of external forces. The four groups may be defined as follows.
An Italian perspective of Chinese specialized towns
121
The ‘Generic’ Manufacturing Bases The first group is composed of a set of 16 specialized towns which show a relatively low level of specific manufacturing specialization, a relatively scarce development of the local tertiary industry, and a limited weight of local specialized markets. These characters suggest that the group may be composed of assembling platforms for large foreign17 or Chinese companies operating in various, possibly non-related, sectors.18 Here local SMEs are often inserted within quasi-hierarchical value chains, extending on a national or global scale, where extra-local clients and buyers specify what, how and how much has to be produced (Humphrey and Schmitz, 2002). The upgrading and development of more sophisticated and autonomous competencies among local SMEs are possible, but they require strong policy efforts. However, empirical evidence on those kinds of upgrading processes is still limited,19 and policy action seems still strongly focused upon just keeping external investors in loco. The Inward-oriented Areas This group of 22 specialized towns is more characterized by the specialized industry than the previous one. Local collective action aimed at realizing specialized infrastructure supporting trade or innovation does not seem (or perhaps does not seem yet) specialized.20 The presence of foreign companies is relatively weak. The group includes a number of ‘traditional’ areas (Long, 2005): many of them are specialized in agriculture-related products, while others are specialized in textiles, garments or footwear production. According to our fieldwork information, local enterprises (mostly local-owned SMEs enterprises) characterize the clusters, producing low to medium quality goods mainly directed to regional or sometimes national markets. The Proto District Areas This group includes 23 specialized towns where the activity of the localowned enterprises is strongly focused on the specialized industry. The presence of a rich tertiary sector and a variety of infrastructure for the enterprises (specialized markets, exhibition centres, and so on) suggests that the local enterprises, mostly SMEs, may find in loco various connections with external markets. Foreign firms do not seem to have a clearly dominant role, as our fieldwork research confirms. Local collective action supports the creation of innovation centres, often as a consequence of the special policy promoted
122
SMEs in a globalised world
by the DST.21 The group includes the most famous specialized towns, such as Nanzhuang, Dachong, Xiqiao, Shiling and Pengjiang, the majority of which are specialized in products for the person and for the house. According to our fieldwork research, the industrial clusters have roots in a local trade or craft tradition, and they are mainly based on the activity of local-owned enterprises, mostly SMEs, producing medium quality products, often with their own brand which are relatively well-known at regional or even national level. Some of these clusters are inserted within global value chains. The combination of local specialized competencies and collective infrastructure supporting the enterprises in entering the markets gives the local SMEs a certain degree of autonomy in the relationship with foreign buyers. This suggests that market or networked forms of interaction with foreign clients are more common than quasi-hierarchical forms. The Foreign-driven Growth Poles The last group is a small group (five specialized towns) that differs from the others for the (relatively) high incidence of the foreign enterprises. The group includes the most developed areas of the province (both in terms of GDP per capita and of industrial output value), localized in the Dongguan prefecture, adjacent to the special economic zone of Shenzhen and close to Hong Kong. Local policies are basically oriented to the attraction of foreign investors, as exemplified by the importance of industrial parks. The high incidence of the migrant workers characterizes this group. The pool of cheap labour fuels the productive capacity of the foreign enterprises (and, obviously, also of local-owned ones). This combines with a relatively strong development of the tertiary sector and the presence of local specialized markets, while collective services for innovation in the local SMEs tend to be weak. Fieldwork research confirms that the industrial clusters are here fully included within global value chains, on the basis mainly of quasi-hierarchical relations. The foreign clients are part of the clusters and they often exert a strong control over the production of the local SMEs. It is true that such strong relations may also facilitate the transfer of technical knowledge and business practices, therefore expanding the possibilities of SMEs upgrading. However, at the moment, case studies provide a limited evidence of success stories (Enright et al., 2005). The results of the multivariate analysis discussed here offer a framework for deeper investigation into the Chinese models of local development. A closer examination of some cases helps in specifying the different kinds of competitive challenge represented by Chinese new industries to European SMEs clusters, like those featuring in Italian industrial districts.
An Italian perspective of Chinese specialized towns
123
THREE CASE STUDIES An example of the many differences between Chinese local industrial bases is illustrated by Caloffi and Hirsch (2004) in their comparison of three industrial clusters belonging to the fashion industry in Guangdong.22 In two of these three cases, the fashion industry is the main industry of a specialized town, that is, Pingzhou and Xiqiao (Nanhai District, Foshan Prefecture), while in the third case it is only one of the various industries located in the territory of the biggest city between Guangzhou (Canton) and Shenzhen, that is Dongguan city. Xiqiao offers an example of a specialized town belonging to the ‘District’ group, while the industrial cluster of Dongguan may be considered as an example of the foreign-like model of development, even if it has not been included in the previous analysis, since Dongguan is not a specialized town but a big city. Pingzhou (hosting a footwear cluster) is an example of an ‘Inward’ oriented area, even if in transition.23 The Xiqiao industrial cluster shows characteristics of embeddedness in the local context, with a handcraft tradition in textile manufacturing dating back to the Ming Dynasty. At present, the large majority of the local enterprises are SMEs owned by local entrepreneurs, and many of them produce and sell their products with their own brand, mostly on the national market. Most of the local enterprises operate within the local specialized market, a large area created by the local and provincial governments which hosts several shops and a logistic centre. The local innovation centre, created in 2000, provides some basic services in design, quality testing and technological assistance. Private Chinese-owned enterprises represent the majority also in Pingzhou. They produce mainly medium-low quality leather shoes, and to a lesser extent also sport shoes, mostly producing as subcontractors for the larger companies run by Taiwanese entrepreneurs. Around half of the total production of the cluster (in volume terms, for the year 2003) is exported. The main export markets are Taiwan, Hong Kong, other South East Asian countries, Russia and Africa. In the Dongguan footwear industry the Taiwanese presence is considerable too; however what mostly characterizes the cluster is the strong presence of large multinational companies. The local footwear enterprises in Dongguan are private and work as subcontractors for foreign firms not only for Hong Kong and Taiwanese companies but also for well-known international brands. Although SMEs form the majority of these subcontractors large local firms are also growing. The heavy presence of migrant workers fuels the productive capacity of foreign enterprises. Although Dongguan is one of the youngest Guangdong shoe manufacturing bases,
124
SMEs in a globalised world
it has today a leading position in export activities, and it is dominating China’s exports of relatively high-quality shoes, followed by Shanghai. The exhibition centres, including footwear and shoe material trade fairs, have grown rapidly in the last decade, and today they rank high in China. The activity of the firms, particularly in Dongguan and Pingzhou, is focused on the assembling of shoes, following a design provided by the clients. Even the materials are often provided by the client. To a lesser extent this is also the case for the technology. The internal organization of SMEs is simple in both clusters. The main difference regards the type of clients the enterprises produce for, a factor that influences the quality level of the product, and obviously the export markets. As Dongguan footwear firms produce for renowned brands, the product is often of higher quality and it is sold in a vast range of international markets. The linkages with foreign enterprises may also positively influence the level of technology of the Chinese enterprises, the international clients being able to introduce – transfer new technologies, machinery and knowledge (Caloffi and Hirsch, 2004). So, the capacities to produce medium-quality goods at a low cost and to sell them at a good price on international markets are under the strategic control of those international companies (Bailey et al., 1999; Schmitz, 2004). Their organization allows the combination of standardized manufacturing in China with technological innovation, differentiated design, and top class marketing on global markets, undertaken in Europe or in the USA. Such international capacities are complemented in the Chinese clusters by the growth of local entrepreneurs. Working as subcontractors for the large international firms, some of them learn basic management and marketing skills. Therefore they develop some local industrial and market capacity independent from the orders of the international firms, even if based on imitation at least at the beginning. Coming to a district-like area such as Xiqiao, it is to be pointed out that, particularly from the mid-1990s, the local government activity has been focused on promoting innovation in the local industry with a broad range of services for the specialized enterprises, such as the aforementioned innovation centre, but also through funding programmes for the renewal of the machinery and the diffusion of CAD technologies. Moreover it has acted for the constitution of a broader group of local public services (such as schools, nurseries and medical clinics), also supporting local development. In the case of Pingzhou, as a comparison, the supportive policy action is more limited. In 2004 the local government propped up the realization of a local innovation centre, but its activity never really took off.24 At present, it is administrated by a private Chinese enterprise, the biggest shoemaking firm located in Nanhai.
An Italian perspective of Chinese specialized towns
125
The case of Xiqiao confirms the identification of the forces at work in what we have called ‘Proto-district’ areas. The jump over a higher road of development, not dissimilar to those experimented by the nascent Italian industrial districts, has some possibilities to succeed, given the density and tradition of local entrepreneurship and the focused action by the local and provincial governments.
CHINESE CHALLENGES TO ITALIAN TYPEDISTRICTS AND WAYS OF REACTION At a general level, and considering recent Italian debates,25 it is plausible to say that the extent of the Chinese challenge is felt in various changing ways within and among Italian industrial districts. The growth of productive capacity and exports from China in many classes of goods for the person and for the house, where the ‘made in Italy’ label and Italian districts have had an international leadership from the 1970s, was very strong in the 1990s. At the beginning of the new decade China had reached or even surpassed Italy in terms of exports (value) in world markets in these fields. For example in 2000 Italy was second and China was first in the world rank of exporters of male clothing, female clothing, footwear, leather products, wool yarns and silk fabric; Italy was first and China was second as regards eye glasses and frames, and ties and silk garments.26 From 2001 to 2004 exports of ‘made in Italy’ commodities have met strong if not homogeneous difficulties, in particular with many classes of goods for the person and the house registering stagnating or falling export trends (Bellandi and Caloffi, 2006). Such difficulties originated from various conditions. Among them the strengthening of the euro against the dollar has brought about general negative effects. However, the fact that the yuan has not been allowed to rise sufficiently against the dollar has made things worse in terms of competitive challenges coming from China, at least for some classes of Italian producers. In general, competitive threats have not been impending for districts and producers positioned on high quality products, where the mix of intangible services, creativity and technology is not easily transferable nor imitable. This is true both of highly-specialized mechanical goods and instruments, where some Italian districts have a strong leadership too, and of the more customized and top tier goods for the person and the house (Guelpa and Micelli, 2007). However, the threat has been real for large parts of the low-to-medium quality manufacturing producers, especially for several sets of smaller industrial firms and craft producers. Bellandi and Di Tommaso (2006) include various examples illustrating how the
126
SMEs in a globalised world
threat in these cases comes from the existence of productive capacities localized in China for manufacturing long series of standardized products at very low costs. The analysis in the previous sections allows, among other results, to give a clearer view of the sources and places of such productive capacities. Their capacities reside quite naturally within clusters of the type found in the ‘Generic’ and in the ‘Foreign Driven’ bases, both within specialized towns and in the industrial parks of special economic zones of large cities. The threat has a fundamental pillar in the FDI and in the buying activities carried out by large US and European companies, including Italian ones, operating as leading actors within hierarchical or quasi-hierarchical global value chains. Medium-quality products come from China to the shelves of big retailers in American and European markets, showing the corporate brands of the large manufacturing companies or of the retailers themselves. Sometimes they come with ‘made in Italy’ or ‘stylized in Italy’ labels if they include some input from Italian productive and design capacities. Chinese products with fake labels and brands are also diffused, generally outside the networks of big retailers. It seems plausible that they come mainly from the development of the independent industrial capacities of Chinese subcontractors, like those in Dongguan, combined with the action of western traders. A more dangerous threat is perhaps at the horizon, given the combination of international value chains and the development of local capacities for quality and innovation in ‘Foreign driven’ bases, and especially in evolving ‘Proto-districts’. The industrial and trade reaction against the Chinese challenge, aimed at increasing the competitive strength of Italian districts, their SMEs clusters, and district-like systems in other old industrial countries, includes business strategies and related institutional support along the lines present in terms of both general and specific characters. Three general lines are recalled normally within the debates on globalization and challenges to Italian districts: (a) strengthening the local capacity to produce and sell products of high and medium quality, and promoting the association with intangible aspects or services difficult to transfer; (b) building local capacities of international management for realizing or purchasing some parts of final and intermediate standard goods in other more specialized localities and clusters, while preserving the control of the filière; and (c) reaching scale thresholds in trade and distribution channels, and getting directly to international markets, partly reducing the dependence from large traders and large manufacturing firms. Lines of reaction specific to China, apart from the invocation of trade protection against dumping and counterfeit products, concern demand
An Italian perspective of Chinese specialized towns
127
and supply opportunities. Chinese markets may be huge and growing, but they are segmented and quite expensive to reach – with geographical, institutional and cultural distance representing a high barrier for SMEs. This normally demands quite heavy investments in trade or industrial facilities, realized directly in China; the constitution and maintenance of joint ventures and institutional relations with Chinese private and public counterparts; the strengthening of the local capacities for innovation and quality. In other words, within China, export channels need to be supported by a stable presence in place, through international investments. A similar type of problems concerns the access to the capacities of Chinese producers to supply components, semi-finished products and final goods, if something different from very standard or low quality products is looked for. The possibility for district companies to make such types of international investments may be questioned. However, the presence of some district characters in a number of Chinese specialized towns suggests a way forward. We defer to other publications a deeper discussion of this point (Bellandi and Caloffi, 2008). Just to give a clue, let us suppose that a Chinese specialized town has the potentiality to develop district processes (like possibly in ‘Proto-district’ areas), and that links with the competencies and products of one Italian district be identified. Then a set of interrelated investments may be driven by a district medium-sized firm and/ or by a district team of smaller but dynamic firms. The set creates a field of stronger quasi-district relations that diffuses and expands in the propitious environment of a specialized town which is a proto-district. Other joint trade or production projects tend to emerge. The risk of transferring strategic components of the ‘district technology’ has to be considered too. In fact, these strategies have to be put in place, from the home district together with the strengthening of local capacities to innovate, so as to combine manufacturing skillfulness and intangible services. Adapted forms of institutional support and collective action at various levels are needed, in particular for building appropriate specific public goods, both at the local and trans-local levels.
CONCLUSION The bases for the specific competitive challenge brought about by Chinese new industries to SME production systems, districts and the like, in Italy and elsewhere, include various combinations of strong public policies, international business strategies, and local forces. They are clearly visible at the level of the many cases of specialized towns and clustered industries scattered in the booming Chinese regions. The challenge is strong, even in
128
SMEs in a globalised world
prospect, since the same bases support a constant upgrading of capacities, some of them similar to those that have featured the development of the Italian districts. This chapter has proposed a closer examination at some aspects of the challenge, through the illustration and the results of a statistical analysis on a dataset of Chinese specialized towns. The analysis and the interpretation have been driven by the reference to models of local development, industrial districts and SMEs clusters. The analysis could be improved in various ways, with the possible acquisition of new data, for example on specific characters of the firms at the level of towns. However, the results already appear to be quite consistent and supported by fieldwork experience. They could be summarized in four points. First of all, the application of a statistical analysis to Chinese new industries at the level of specialized towns gives a robust confirmation of the existence of various forms of industrial development in Chinese specialized towns, in Guangdong in particular; the understanding of such forms may be helped by the reference to European models of local development, industrial districts and SMEs clusters. Second, the challenge to Italian districts and the like may come both from forms showing features distant from the models above, that is, the so-called foreign driven industrial bases, and from forms nearer to the model, that is, the so-called proto-districts. Third, these forms show consistent features, confirming the strength of the challenge, even in prospect, but they also give clues for strategies of international investments promoted by teams of local SMEs and public agents, and aimed at finding partnerships within Chinese new industries. Finally, the internationalization strategies of SMEs may be helped when they operate within district and cluster contexts; nonetheless even the SMEs accustomed to trade relations with foreign markets and operating within favourable contexts are asked difficult adaptations in order to incorporate and implement such strategies in a successful way.
NOTES 1.
2.
Marco Bellandi is Professor of Applied Economics, Faculty of Economics, University of Florence. Annalisa Caloffi is Doctor of Research, University of Florence, Sections 3, 4, 5 have been written specifically by Annalisa Caloffi; sections 2, 6, 7 by Marco Bellandi. The authors wish to thank all the participants in The China and Italy Research and Learning Project, a project promoted by a web of Italian and Chinese partners and coordinated by Professor Marco Di Tommaso; Professor Marco Bellandi; and Professor Long Zhihe. The China and Italy Research and Learning Project has a follow-up as a line of activity of c-MET 05 (www.cmet.it). Previous versions of this
An Italian perspective of Chinese specialized towns
3. 4. 5.
6.
7. 8. 9. 10.
11.
12.
13. 14.
15.
129
chapter have been presented at the 9th EUNIP International Conference, Department of Economics, Kemmy Business School, University of Limerick, Ireland, 20–22 June 2006, and at the International Conference on Regional Integration – Asia and Europe Compared, Università Ca’ Foscari, Venice, 16–17 January 2008. Grateful acknowledgements extend to the organizers and audiences of these meetings and to the editors of this book. What is written in this chapter is of course the sole responsibility of the authors. The international negotiation of transnational or global rules of interaction is of course an important element. These three points here also represent the main interpretative results elaborated within the ‘China and Italy Research and Learning Project’ referred to in Note 2. As stated by the Provincial government, in order to receive the official designation, the specialized town (ST) has to satisfy the following requirements: (a) from an administrative point of view it has to be a town, a county or an urban district; (b) in terms of sectoral specialization, at least 30 per cent of the manufacturing output (and/or employment) has to be produced by the specialized industry; and (c) the annual industrial output value of the specialized industry has to be more than 2 billion Yuan. For more information on aggregate measures concerning a core sample of specialized towns, see Bellandi and Di Tommaso (2005). The economy of the two deltas is characterized by a variety of big cities playing as gateways for national and international trade and investment, a rich set of localities where various types of productive activities are clustered within a strong regional infrastructure including transport, communication, universities, and science and technology parks (Enright et al., 2005). For example, from multinational enterprise (MNEs) – see Part 2 of this book. Cooke (2002, p. 127) writes on the last point, referring a sensible use of the term ‘cluster’ to a rich meaning, tied to the idea of local development. Shanghai and Sunan together with the northern areas of Zhejiang correspond to the Yangtze River Delta proper – with Wenzhou being located more southward. Actually some authors refer to different models of Chinese industrialization: the Wenzhou model, the Sunan model, the Dongguan model. They could be thought of as the extremes of the triangle of factors behind Chinese new industries referred to before (local forces, public policies, international business) and combined in various ways with the access to cheap and disciplined labour, cheap land, improving infrastructure, large and increasing internal markets. See Di Tommaso and Rubini (2006) and the information within the Acknowledgement section. Following Qiu and Xu (2004), some 122 specialized towns were identified in Guangdong Province at the beginning of this decade, and some of them (among those officially acknowledged) applied to take part to a special programme implemented by the provincial government in 2003, and aimed at supporting the creation of innovation centres. These data refer specifically to the administrative boundaries of the principal town (or the urban district) located at the core of the specialized town, while the specialized area may also extend to a set of nearby smaller villages, rural semi-urbanized areas localized in nearby towns. It follows that the quantitative individual and aggregate importance of this set of Guangdong Specialized towns is larger when the wider territorial units are considered. The fieldwork investigation has been carried out by Annalisa Caloffi – together with the participants to the China and Italy Learning and Research Project – in April–June 2004 and April–July 2005, with the help of the South China University of Technology. The official statistics (Fifth Population Census) define the migrant (or floating) population as the ‘population whose census registration place is different from the household registration place and whose household registration place is under the jurisdiction of other counties, county-level cities or cities at prefecture level in or out of a Province’ (NBS 2004b, p. 82) See for example Qiu and Xu (2004); Caloffi and Hirsch (2004).
130 16.
17. 18.
19.
20. 21. 22. 23.
24.
25. 26.
SMEs in a globalised world At a first stage, in order to identify the ‘optimal’ number of clusters, we have used a hierarchical agglomerative algorithm. Four clusters have been identified on the basis of the observation of the Calinski and Harabasz (1974) value (as provided by Stata software) for cluster solutions and of the cluster tree. At a second stage, we have used a k-means technique in order to ameliorate the composition of the four clusters previously identified (see Figure 6.1). A first version of this analysis was elaborated by Caloffi (2005). The group exhibits the lowest distance with the foreign-driven areas (Table 6.2). The specialized town of Shijie – belonging to this group – offers a typical example. This specialized town has developed around the presence of a foreign (Taiwanese) enterprise, which directly employs about half of the local specialized employees and which organizes the activity of several (dependent) assembling enterprises, providing them with components and machinery. The group includes also specialized towns whose activity is not so dependent from that of foreign enterprises. It is the case of Nantou, where a set of Chinese large enterprises specialized in the production of electrical household appliances has their manufacturing bases (Di Tommaso and Rubini, 2006). This seems to be happening in Shijie, where some ex-managers (mostly of Taiwanese origin) of the leader enterprise have founded autonomous enterprises (data collected by fieldwork investigation conducted with the help of South China University of Technology – see Note 13). The lack of local collective infrastructure – which often parallels the lack of such infrastructure at prefecture, city and county level – does not seem to be counterbalanced by the development of capacities internal to single local companies. An index such as INNOCEN is connected to relevant forms of local collective action (Section 3). A ‘district’ group similar (even if smaller) to the previous one can be detected even when the cluster analysis is performed without that index. The conditions illustrated refer to the years 2003–04. According to Zhongshan University (2004): (a) in Xiqiao cluster, 1670 private enterprises specialized in textile and clothes made by artificial fibres, 60 000 employees that is, around half of the total workforce of the town; (b) in the Pingzhou cluster, 600 specialized enterprises, with 100 employees on average; (c) in the Dongguang cluster, 1200 specialized enterprises, with 250 employees on average. The establishment of the centre resulted from a planned activity, not supported by a preliminary market research and without any attempt to involve the local enterprises. Not surprisingly the centre suffered a problem of demand, as the enterprises often did not perceive the importance of the services offered by the centre. Moreover in some cases the staff of the centre were not skilled enough in offering qualified services, nor they were able to identify appropriate external consultants (Caloffi and Hirsch, 2004). See for example various chapters in Cainelli and Zoboli (2004) and in Bellandi and Di Tommaso (2005). Data and ranks from Tradstat (www.tradstat.com) refer to the year 2000 and are reported in Quadrio Curzio and Fortis (2003).
REFERENCES Arndt, Sven W. and Henryk Kierzkowski (eds) (2001), Fragmentation: New Production Patterns in the World Economy, Oxford: Oxford University Press. Bailey, David, Gary Harte and Roger Sugden (1999), ‘Regulating transnationals: free markets and monitoring in Europe’, in Keith Cowling (ed.), Industrial Policy in Europe, London: Routledge, pp. 311–25. Becattini, Giacomo (2004), Industrial Districts: A New Approach to Industrial Change, Cheltenham, UK and Northampton, MA, USA: Edward Elgar.
An Italian perspective of Chinese specialized towns
131
Bellandi, Marco and Annalisa Caloffi (2006), ‘Distretti industriali italiani e internazionalizzazione fra gli anni novanta e la prima metà del nuovo decennio’, in Rapporto ICE 2005 – 2006, L’Italia nell’economia internazionale, Roma: ICE, pp. 466–82. Bellandi, Marco and Annalisa Caloffi (2008), ‘District internationalization and translocal development’, Entrepreneurship and Regional Development, 20 (6), 517–32. Bellandi, Marco and Marco R. Di Tommaso (2005), ‘The case of specialized towns in Guangdong, China’, European Planning Studies, 13 (5), 707–29. Bellandi, Marco and Marco R. Di Tommaso (2006), ‘The local dimensions of industrial policy’, in Patrizio Bianchi and Sandrine Labory (eds), International Handbook of Industrial Policy, Cheltenham, UK and Northampton, MA, USA: Edward Elgar. Biggeri, Mario, Danilo Gambelli and Christine Phillips (1999), ‘Small and medium enterprise theory: evidence for Chinese TVEs’, Journal of International Development, 11 (2), 197–219. Cainelli, Giulio and Roberto Zoboli (eds) (2004), The Evolution of Industrial Districts. Changing Governance, Innovation and Internationalisation of Local Capitalism in Italy, Heidelberg: Physica-Verlag. Calinski, Tadeusz and Jutman Harabasz (1974), ‘A dendrite method for cluster analysis’, Communication in Statistics, 3, 1–27. Caloffi, Annalisa (2005), ‘Le politiche per l’innovazione e l’internazionalizzazione in ottica interdistrettuale. Un’applicazione al confronto fra sistemi di produzione locale italiani e cinesi’, PhD dissertation at the University of Florence, Florence, Italy. Caloffi, Annalisa and Giovanna Hirsch (2004), ‘Cluster-to-cluster analysis: zooming on Guangdong fashion industries’, in Marco R. Di Tommaso (ed.), The China and Italy Joint Research and Learning Project: Clusters, Industry and Industrial Policy, Farrara, Italy: University of Ferrara, pp. 356–98. Chelan Li, Linda (2001), ‘Guangdong playing the “foreign card”: politics and economics across territorial boundaries’, in Glenn Drover, Graham Johnson and Julia L.P. Tao (eds), Regionalism and Subregionalism in East Asia: The Dynamics of China, Huntington, New York: Nova Science Publishers, pp. 151–67. Cheng, Joseph (1998), The Guangdong Development Model and Its Challenges, Hong Kong: City University of Hong Kong. Christerson, Brad and Constance Lever Tracy (1997), ‘The third China? Emerging industrial districts in rural China’, International Journal of Urban and Regional Research, 21 (4), 569–88. Cooke, Philip (2002), Knowledge Economies. Clusters, Learning and Cooperative Advantage, London: Routledge. Di Tommaso, Marco R. and Lauretta Rubini (2006), ‘Cluster industriali e specialized towns del Guangdong: la centralità del Delta del Fiume delle Perle’, in Marco R. Di Tommaso and Marco Bellandi (eds), Il fiume delle perle. La dimensione locale dello sviluppo industriale cinese e il confronto con l’Italia, Torino, Italy: Rosenberg and Sellier, pp. 183–211. Eng, Irene (1997), ‘The rise of manufacturing towns: externally driven industrialisation and urban development in the Pearl River delta of China’, International Journal of Urban and Regional Research, 21 (4), 554–68. Enright, Michael J., Edith Scott and Ka-mun Chang (2005), Regional Powerhouse: The Greater Pearl River Delta and the Rise of China, Singapore: John Wiley and Sons.
132
SMEs in a globalised world
Gereffi, Gary and Miguel Korzeniewicz (eds) (1994), Commodity Chains and Global Capitalism, Westport, CT: Praeger. Guelpa, Fabrizio and Stefano Micelli (eds) (2007), I distretti industriali del terzo millennio. Dalle economie di agglomerazione alle strategie di impresa, Bologna, Italy: Il Mulino. Humphrey, John and Hubert Schmitz (2002), ‘How does insertion in global value chains affect upgrading in industrial clusters?’, Regional Studies, 36 (9), 1017–27. Long, Zhihe (2005), ‘A study on the industrial clusters in Guangdong’, paper presented at the international conference Enter the Dragon: China’s Emergence and International Competitiveness, 7–11 November, Hong Kong. Ma, Laurence J.C. and Chusheng Lin (1993), ‘Development of towns in China: a case study of Guangdong Province’, Population and Development Review, 19 (3), 583–606. National Bureau of Statistics (NBS) (2004a), China Township and Village Statistics, Beijing: China Statistics Press, in Chinese. NBS (2004b), Guangdong Statistical Yearbook 2003, Beijing: China Statistics Press. Qiu, Haixiong and Jianniu Xu (2004), ‘The actions of local governments in the technological innovation of industrial clusters’, paper presented at the international conference Regional Innovation Systems and Science and Technology Policies in Emerging Economies: Experiences from China and the World, 19–21 April, Guangzhou, Zhongshan University. Quadrio Curzio, Alberto and Marco Fortis (2003), ‘Un marchio d’origine contro la contraffazione’, Economia e politica industriale, 118, pp. 21–5. Schmitz, Hubert (2004), Local Enterprises in the Global Economy: Issue of Governance and Upgrading, Cheltenham, UK and Northampton, MA, USA: Edward Elgar. South China University of Technology (2005), ‘A survey on the furniture produce enterprise cluster in Longjian Town of Foshan City’, mimeo. Walcott, Susan M. (2003), Chinese Science and Technology Industrial Parks, Aldershot, and Burlington, VT: Ashgate. Wang, J. (2006), ‘The general tendency of China’s industrial clusters in the globallocal nexus’, presentation to the International Workshop on Asian Industrial Clusters, 29 November–1 December, Lyon. Wang, Jici and Xin Tong (2005), ‘Industrial clusters in China: embedded or disembedded?’, in Claes G. Alvstam and Eike W. Shamp (eds), Linking Industries Across the World: Processes of Global Networking, Aldershot and Burlington, VT: Ashgate, pp. 223–41. Yao, Yang (2001), ‘Social exclusion and economic discrimination: the status of migrants in China’s coastal rural areas’, China Center for Economic Research working paper, no. E2001005. Yeung, G. (2002) ‘WTO accession, the changing competitiveness of foreignfinanced firms and regional development in Guangdong of Southern China’, Regional Studies, 36 (6), 627–42. Zhongshan University (2004), ‘A case study of the textile cluster of Xi Qiao, Nan Hai – an analysis of the technological innovations’, mimeo.
PART II
MNEs, SMEs and industrial development
7.
MNE subsidiaries, productivity spillovers and SMEs Rita Buckley
INTRODUCTION Prior to 1990, the Irish economy was considered to be one of the weakest peripheral economies of the ‘old’ EU15 with per capita GDP below 80 per cent of the EU average. This was combined with slow growth rate in GDP, high inflation and unemployment and a heavy debt burden. In 2007, the Irish economy was among the most vibrant, globalized economies in the EU, with GDP per capita second only to Luxembourg’s at 143 per cent of the EU average (Eurostat, 2008). The ‘Celtic Tiger’ period of the mid-tolate 1990s saw several years of double-digit GDP growth, driven by a progressive industrial policy that boosted large-scale foreign direct investment and exports. Since 2004, GDP growth has averaged around 5 per cent, the best performance for this period among the original EU15 Member States. This extraordinary growth record of the Irish economy has led to considerable interest in the Irish development model. Ireland’s economic performance since 1990 is primarily attributed to the presence of foreign-owned multinational enterprise (MNE) subsidiaries, especially in high-technology sectors. This sector is made up of a relatively small number of large foreign-owned multinational enterprises and a large number of indigenous small- to medium-sized enterprises (SMEs). A key issue here is how the indigenous SME sector has coped with the challenge of coexisting with a highly productive ‘foreign’ sector and whether this has had an impact on their overall level of growth and in particular whether it has assisted them in coping better with the pressures of globalization and if so, how exactly and to what extent can their performance be linked? There is already evidence to suggest that software SMEs have responded well to the challenges of globalization in Ireland. Empirical evidence suggests that MNEs have encouraged new firm creation in the manufacturing and software sectors (Görg and Strobl, 2002; Giarratana et al., 2003). In the Irish indigenous software industry, the majority of new firms have been created by entrepreneurs that had previous work experience in foreign 135
136
SMEs in a globalised world
software and hardware companies (Buckley, 2005). This may indicate that spillovers in knowledge carried by former employees of software MNEs are applied to the ‘new’ domestic firm which makes them more productive relative to other domestic firms in the same industry. These spillovers carried by new owners through worker mobility in software represent the knowledge and experience gained by an employee from past employment in a MNE, that has some applicability to production, management practices and techniques and R&D utilization in the domestic SME. If this is the case, indigenous SMEs in the Irish software sector might be significantly benefiting from the presence of foreign software MNEs. The question of whether foreign-owned MNE subsidiaries in Ireland has contributed to productivity improvements in domestic SMEs can shed more light on this issue. This chapter presents new evidence on the impact of MNE subsidiaries on the productivity performance of SMEs using the Irish software industry as a case study. The potential for productivity spillovers from foreign affiliates to domestic firms is an important rationale for host country industrial policies focused on the attraction of MNEs. It is well known that MNE subsidiaries can directly benefit the host country especially with respect to employment creation and export growth. However, governments have now recognized the potential for substantial indirect benefits from the presence of foreign MNE subsidiaries in the host economy. Such indirect benefits or productivity spillovers include the transfer of technology, the training of workers and the development of backward and forward linkages between the domestic and foreign sector. The following sections empirically investigate the impact of MNE subsidiaries on the productivity performance of small firms in the Irish software industry. The indigenous software sector in Ireland has grown rapidly since the early 1990s and is dominated by SMEs. This rapid growth occurred in tandem with significant increases in the number and size of MNE software industries in Ireland. However, until now, the connection between MNE entry and the rapid development of the indigenous software industry has not yet been empirically explored. The analysis contained in the following sections allows us to draw important lessons regarding the extent to which SMEs benefit from productivity spillovers and the degree to which the domestic SME sector can coexist with the presence of large foreign-owned MNE subsidiaries.
MNEs AND PRODUCTIVITY SPILLOVERS The term ‘spillover’ is generally used in the MNE literature as a catch-all term designed to define the perceived residual benefits that accrue to host
MNE subsidiaries, productivity spillovers and SMEs
137
country firms as a result of foreign entry (Harris and Robinson, 2002). For spillovers to occur the MNE subsidiary must be technologically superior to domestic firms in the host country. Such technical superiority or firm specific advantage is generally transferred from the parent company to enable the subsidiary to compete effectively with domestic firms and can consist of anything from production technology to international management and marketing skills (Blomström et al., 2000; Görg et al., 2006). Sometimes part of the subsidiaries firm specific advantage is transmitted to domestic firms, which raises the productivity of the indigenous sector (Blomström and Kokko, 2001). The theoretical literature on productivity spillovers has focused on how productivity improvements are diffused from the MNE subsidiary to the local firm or industry in the host country. There are three major channels considered. The first channel transmits productivity improvements to domestic firms through demonstration effects. Backward linkages, where foreign firms demonstrate new technologies and provide technological assistance to their local suppliers and customers are particularly relevant in this context (Görg and Strobl, 2002; Javorcik and Smarzynska, 2004). The second channel operates through competition effects where the presence of foreign-owned firms in an economy increases competition in the domestic market. The competitive pressure may spur local firms to operate more efficiently and introduce new technologies earlier than would otherwise have been the case (Caves, 1996; Görg and Greenway, 2004). The third channel operates through the labour market. Here, productivity benefits can accrue to domestic firms through labour migration between MNE affiliates and host country firms. The productivity gain arises from the enhancement of human capital due to training of labour, especially management in unique skills (Barry et al., 2004; Lipsey and Sjöholm, 2005). Employees who later switch to domestic firms from MNE subsidiaries may transplant knowledge to indigenous firms that provide a basis for imitative productivity gains without incurring any real resource costs (Görg and Strobl, 2002). Since the mid-1980s, there have been numerous empirical studies attempting to evaluate the existence, magnitude and general impact of productivity spillovers from foreign presence on recipient countries. The majority of these studies test for productivity spillovers in the manufacturing sector and have produced, at best, ambiguous results. Contradictory results could be explained by the fact that differences in the existence and magnitude of spillovers vary according to the overall level of development of the host economy (Blömstrom et al., 2000; Javorcik and Smarzynska, 2004). Developing and transition economies are more likely to be dominated by small, weak and technologically backward firms and therefore less
138
SMEs in a globalised world
likely to register positive spillover effects from MNE affiliate activity. In contrast, host economies which exhibit a high level of technological capability and a well-developed industrial sector are likely to benefit more from international production (Javorcik and Smarzynska, 2004; Wei and Liu, 2006). In addition, there may be differences between individual industries in the level of technology and knowledge within the host economy itself, which may explain why the presence of spillovers are noted in some sectors (especially high technology sectors) and not in others (Kokko, 1996; Görg and Greenway, 2004). Other studies have noted that differences in market size, technology, level of protection and availability of skills impact on the magnitude of spillovers (Blömstrom and Kokko, 2001). While yet other studies (for example, Görg and Strobl, 2002; Hale and Long, 2007) relate the ambiguity of the results to the use of different methodologies and conclude that studies using cross-section data are more likely to establish the existence of positive spillover benefits to the host economy.
THE IRISH SOFTWARE INDUSTRY The Irish software sector has experienced rapid growth especially since the early 1990s and makes a significant contribution to the Irish economy. In 2006, software output accounted for over 11 per cent of GDP and over 14 per cent of total exports (ISA, 2007). Importantly, the indigenous Irish software industry has developed rapidly in parallel with the presence of large foreign-owned software companies. Most leading US software vendors such as Microsoft, Oracle and Symantec, have based their European operations in Ireland, especially in the Dublin region. Moreover, Ireland is also the prime location for software localization activities in Europe. However, there are substantial differences between foreign and domestic firms especially, with respect to firm size, exporting behaviour, revenues and productivity. The vast majority of indigenous firms are SMEs whereas foreign firms in this sector are generally quite large. Average employment in indigenous software firms is 14 compared with an average employment of over 100 in foreign software subsidiaries in 2003. Total employment, both indigenous and foreign, increased from around 7500 in 1991 to just over 30 000 in 2006 and the number of indigenous and foreign firms increased from 365 to over 600 in 2006. Indigenous firms account for over 45 per cent of total employment in the Irish software sector. While, the balance of employment between indigenous and foreign firms has remained fairly constant since 1999, indigenous firms are far more numerous than their foreign counterparts. For example, in 2006 there were 140
MNE subsidiaries, productivity spillovers and SMEs
139
foreign firms compared with over 600 indigenous firms in the same year (ISA, 2007). Growth in employment and firm numbers in the sector has accelerated considerably and new indigenous start-ups were the main driving force behind this trend. The indigenous industry is characterized by firms with strong product focus and a strong export orientation. Over 42 per cent of indigenous firms have a significant level of foreign sales with the UK and the USA being the primary export markets for these companies. However, foreign firms are far more export oriented than domestic ones with overseas firms accounting for 92 per cent of total exports for the sector in 2003. Exports as a percentage of revenue remained fairly constant for foreign firms since 1991 at around 98 per cent. However, there has been considerable improvement in the export performance of indigenous firms over the 1991–2003 period with exports as a percentage of revenue increasing from 40.8 per cent in 1991 to over 80 per cent in 2003 (Buckley, 2005). Growth in revenues in the Irish software industry since 1990 is substantial. There was a significant increase in revenue in the overall software industry from over 2.1 million euros in 1991 to 14.8 million euros in 2003. Again, there is a stark contrast between revenue generated by indigenous firms and that generated by foreign-owned firms. For example, almost 91 per cent of total sectoral revenue in 2003 was due to the activities of the overseas sector with Irish-owned software companies responsible for less than 10 per cent (Buckley, 2005). Although indigenous firms are generally less productive than foreign firms, the statistics relating to productivity are not reliable. Productivity in foreign software firms as measured by average revenue per employee is greatly overstated. For example, the average revenue per employee in the foreign sector was over one million euros compared to just over 127 thousand euros for the indigenous sector in 2003 (Buckley, 2005). From these statistics, it would appear that overseas software firms are far more productive than their indigenous counterparts. However, this is not the case, rather foreign software companies are unusual in having extremely high revenue per employee and per firm (Crone, 2002). This is primarily because of the presence of highly visible US package software firms that have chosen Ireland as their base for selling into the EU market and also due to transfer pricing practices of such firms who are keen to maximize the advantage of the low rate of corporate tax (of 12.5 per cent) available in Ireland. However, despite the differences outlined above there is no denying that the Irish software industry, foreign and domestic, thrived for more than a decade. The growth in the domestic sector occurred in tandem with the growth and development of a thriving foreign-owned sector. A key question to be addressed is whether overseas firms had a role to play in
140
SMEs in a globalised world
explaining the rapid growth and development of the indigenous firms in the software sector.
DATA AND METHODOLOGY A major objective of this research is to determine whether the presence of MNE subsidiaries has raised the productivity levels of domestic firms in the Irish software industry. Higher productivity levels as a result of MNE entry would indicate the presence of positive productivity spillovers. There is much anecdotal evidence suggesting a relationship between the presence of foreign software subsidiaries and accelerated growth in the indigenous industry. However, so far, such a relationship has remained untested until now. This section outlines the source of the individual firm data used in the empirical estimations along with the methodology used in determining whether indigenous firms benefited from foreign presence in the software sector in Ireland. Data The study focuses on indigenous software firms operating in Ireland and is based on a sample of 60 indigenous firms. The source of the sampling frame was the National Informatics Directorate (NID) database which tracks the development of indigenous firms on an annual basis and provides data on firm numbers, employment, turnover and exports for the indigenous software industry since 1991. Using this source, an initial list of firms in the Irish indigenous software industry was compiled from a population of 750 indigenous firms in 2002. A total of 121 firms were randomly selected from this population within the sampling framework outlined below. All companies were contacted initially by email and then telephone and were invited to participate in the study. Structured interviews based on a questionnaire were conducted with Chief Executive Officers (CEOs) or owners of 69 of these software companies over the summer and autumn of 2003. Nine of these companies were subsequently dropped because they did not meet the criteria or the data collected was inadequate for the requirements of the study. There were a number of firms discarded from the general population of domestic software firms for the purposes of the fieldwork. First, 101 firms from the population of firms were unidentifiable and classified as ‘unknown’. This is due mainly to information deficiencies regarding their key characteristics and/or because they may have ceased operations.1 Second, firms involved in general software/IT training, data entry and
MNE subsidiaries, productivity spillovers and SMEs
141
processing and firms specializing in selling software products for other companies were omitted from the population because of their limited value added for the sector. Finally, firms not set up and operational prior to 1999 were also excluded from the study. This is because productivity improvements from MNE presence to domestic firms in the software sector are not instantaneous. It is only reasonable to expect a time lag between a domestic firm start-up and possible productivity spillovers from overseas MNE subsidiaries (Scott-Kennel and Enderwick, 2006). The study must take account of this if it is to offer an accurate picture of these effects. The final sample coverage of 60 firms resembles relatively closely that of the underlying population in terms of geographical spread, firm size and product and service orientation of the indigenous software industry in Ireland. The sample reflects well the geographical spread of the industry. Over 75 per cent of the population of indigenous firms are located in the greater Dublin area. The Cork and Limerick area accounts for another 9 per cent each of the population with the final 6 per cent of firms located in the Galway region. The geographical distribution of sampled firms is representative of these trends with over 68 per cent of the sample taken from the Dublin region, almost 12 per cent each from the Limerick and Galway regions and over 8 per cent from the Cork region. Numbers of companies were selected by size class in accordance with the proportion of the sectors employment in each size class. Firms in each of the employment categories are included and due weight given to firms especially in the 11–50 employees category which makes up the most significant portion of the population. Firms with greater than 100 employees are excluded from the sample.2 These account for only 1 per cent of the target population and should not bias the results. The sample is also representative of the division between product and service orientation of the indigenous Irish software industry as categorized by O’Gorman et al. (1997). The NID estimated that nearly 60 per cent of indigenous software companies could be classified as product orientated. Fifty-eight per cent of the sample was drawn from such firms to reflect this activity division in the domestic industry. Methodology Almost all empirical studies of productivity spillovers in the general literature test for productivity spillovers in the manufacturing sector. The most standard approach in testing for productivity spillovers in this sector is to regress total factor or labour productivity of domestic firms on a number of independent variables including a measure of the extent of multinational
142
SMEs in a globalised world
presence in an industry to account for the spillover effect. In particular, the majority of empirical studies follow Kokko (1996) and hypothesize that the labour productivity of indigenous firms in manufacturing sectors can be estimated by the following productivity equation:3 (LP) it 5 f (LQit, Kit, HERFjt, FORjt)
(7.1)
Where i is an individual firm/plant, j is a sector and t is an individual year. Labour productivity (LPit) is estimated as a function of labour quality (LQit), capital intensity (Kit), a sectoral level of producer concentration (HERFjt) and the level of foreign presence within the sector (FORjt). The share of all industry employment in foreign-owned plants is the measure of foreign presence used. A positive and statistically significant coefficient on the latter variable is interpreted to indicate the existence of positive productivity spillovers from MNE subsidiaries to indigenous firms. In Kokko’s (1996) model the amount of technology that could spillover from MNE subsidiaries to domestic firms in the manufacturing sector is not exogenously given but dependent on both host country and industry characteristics. The control variables used to account for these effects are labour quality (LQ), capital intensity (K) and degree of concentration (HERF). The model to be estimated is loosely based on that of Kokko (1996) but adapted to take into account the unique characteristics of the software industry, especially its human capital intensity and importance of research and development expenditures for product development. The general empirical specification of the model to be estimated is given by Equation 7.2. The objective of the econometric analysis is to test whether the increasing presence of MNE subsidiaries in Ireland’s software industry contributes positively to the labour productivity of Irish indigenous software firms. LPit 5 b0 1 b1FPjt 1 b2LQit 1 b3RDit 1 b4FSit 1 Tt1 1 uit
(7.2)
where [i = 1,. . .. . ..N; t = 1,. . .. . .T] and [uit = μi + nit] Labour productivity in the indigenous software industry is measured as a function of foreign presence (FP), labour quality (LQ), research and development (RD) and foreign sales (FS). Labour productivity (LPit) Labour productivity is the dependent variable in the model. In previous empirical work on productivity spillovers, value added per employee is
MNE subsidiaries, productivity spillovers and SMEs
143
often used as a proxy for labour productivity (Blomström and Kokko, 2001). However, this research has access to detailed firm-level data, which allows us to use the fairly robust measure of sales divided by employee numbers as the proxy for labour productivity. Labour quality (LQit) Labour quality is of key importance in explaining any variation in the labour productivity of domestic software firms (O’Riain, 2000). The software industry is human capital intensive and relies heavily on the availability of a large pool of qualified workers. This leads us to suspect that the labour quality variable will be highly relevant in explaining labour productivity. Labour quality is proxied by the percentage of firm employees that have at least a third level degree. The expectation is that labour productivity will increase with the quality of human capital employed in the firm. Research and development (RDit) Like many other industries research and development expenditures are important in explaining labour productivity (Coe, 1997). To remain competitive, software firms must consistently support high-levels of R&D. These investments are necessary to develop new products and to upgrade and enhance existing ones (Crone, 2002). R&D expenditures are measured as a percentage of sales. It is expected that labour productivity will increase with additional R&D spending given the importance of new product development and upgrading of existing products. Foreign sales (Fsit) The indigenous software sector is highly export-oriented and depends substantially on revenues from export activity to enhance productivity (Arora et al., 2001). The foreign sales variable is measured as the percentage of firm turnover that is exported. This measures the degree of international orientation of the firm. It is expected that labour productivity may be positively or negatively correlated with exports, depending on conditions in international markets. Foreign presence (FPit) The entry of foreign software subsidiaries is unlikely to have an immediate effect on the productivity of domestic firms. The expectation is that there will be a lag before foreign entry will impact on the labour productivity of domestic firms. The MNE employment share of all software industry employment for each year between 1997 and 2000 inclusive is the measure of foreign presence in the sector. A significant and positive coefficient on the foreign presence variable, after the effects of all other variables have
144
SMEs in a globalised world
been accounted for, is evidence of a productivity spillover effect (Kokko, 1996). Year dummies (Tt) The model also includes full sets of year dummy variables (Tt) for the period 2000 to 2003 to account for macroeconomic changes which impact on the software industry. Year dummies are especially important given the substantial downturn experienced in high technology sectors since 2000. It is expected the year dummies will be inversely related to labour productivity of domestic firms at least for some of these years. Error term (uit) The incorporation of the μi in the error term allows controls for across firm heterogeneity by incorporating a ‘firm effect’ into the error term. As with all panel estimations, the error term consists of two components, μi which denotes the unobservable individual firm specific effect and nit the remainder disturbance. The unobservable firm specific effect, captured by the μi, can be thought of as, for example, entrepreneurial or managerial skills of the firm’s executives which are likely to impact on labour productivity (Baltagi, 2001). Equation 7.1 Equation 7.1 is estimated using firm-level panel data for the sample of 60 firms covering the period 2000–03. Regression results for total Ordinary Least Squares (OLS), fixed effects (FE) and random effects (RE) estimators are presented and the results are compared. Given the panel nature of the data account must be taken of any time invariant firm specific effect. If such effects are present, estimation results may be biased. Both the fixed effects and the random effects model allow for differences across firms whereas the OLS model takes no account of such heterogeneity.4
FINDINGS Foreign Presence and Productivity Levels Table 7.1 presents the regression results for the total OLS, fixed effects (FE) and random effects (RE) regressions. A standard F test fails to reject the hypothesis of no fixed effects which suggest that either the random or fixed effects model is preferred. The Hausman test (HN) statistic helps in the choice between the random effects or fixed effects model. It tests the null hypothesis that the random effects are uncorrelated with the
MNE subsidiaries, productivity spillovers and SMEs
Table 7.1
145
Foreign presence and productivity levels in the Irish indigenous software industry
Dependent variable labour productivity Independent variables RD FS LQ FP Year dummy 2000 Year dummy 2001 Year dummy 2002 Year dummy 2003 Diagnostics R-squared Adjusted R-squared F-test (OLS v FE)a HN (RE v FE)b Condition index DW statistic (p-value) LM test (p-value) No. of obsc N (firms)
Total OLS coefficients
Fixed effects (FE) coefficients
Random effects (RE) coefficients
0.017470 *** (17.289) 0.00004996 (0.205115) 0.00242072*** (5.0841) 1.35197*** (2.63177) −0.009165** (2.15) −0.703970 (−0.08701) −0.43151*** (−6.34123) 0.05793970** (2.65)
0.00217351** (1.88484) 0.00013915 (−0.251574) 0.00162042*** (3.30415) 2.39942*** (7.39240) −0.1584792** (−2.77) −0.156298*** (−7.13443) −0.08108*** (−5.0991) 0.000419 (0.0211)
0.00809206*** (8.42231) 0.0005144273 (1.54890) 0.0027132*** (6.37676) 1.82444*** (6.04451) −0.11992*** (−6.04451) −0.119992*** (−5.79696) −0.063645*** (−4.07344) 0.0036859** (2.20)
0.730489 0.723548 Reject Na 6.23 1.82 0.07813 6.2184 (0.6324) 240 60
0.937317 0.9139012 Favour Favour 2.68 2.063 (0.1632) 5.4065 (0.1783) 240 60
0.682949 0.674785 Na Rejects 1.91 2.314 (0.1361) 7.123 (0.5612) 240 60
Notes: ** significant at the 5 per cent level; ***significant at the 1 per cent level. Figures in parentheses are t-statistics unless otherwise stated. RD is research and development expenditure, FS is foreign sales, FP is foreign presence. a F-test which tests the significance of fixed effects. F(59, 174) = 9.7310. Fcrit = 8.3966. b HN: Hausman statistic for testing the random effects model against the fixed effects model; CHISQ (8) = 87.016, p-value= (0.0000). Rejection of the random effects model in favour of fixed effects. c The number of observations is 240 due to the use of panel data (that is, 4 years of data for each of the 60 firms).
146
SMEs in a globalised world
explanatory variables. The null hypothesis is rejected and hence the model preferred is the fixed effects specification. Nevertheless, all three models are presented for comparative purposes.5 The estimated coefficient on foreign participation (FP) is one that interests us most for the purpose of the research. The positive and significant results reported (at one per cent level) suggests that there is a firm specific productivity gain associated with a high level of foreign employment in the software sector. This result is evidence that productivity spillovers from MNE subsidiaries to domestic firms are present in the Irish software industry. The results for the R&D and labour quality variables coincide with initial expectations and confirm their importance in any explanation of labour productivity in the software industry. Research and development (RD) and labour quality (LQ), all contribute positively to labour productivity and are statistically significant at the 1 per cent level. The initial expectation was that the coefficient on the foreign sales (FS) variable could exert either a positive or negative effect on domestic firm labour productivity. The coefficient on the foreign sales variable is negative and statistically insignificant for the fixed effect model specification. It is likely that the severe downturn in the global high technology industry, especially in the computer software and hardware sector, is the main reason for this result. The coefficients on the year dummies included in the regressions to account for the effect of macroeconomic changes on labour productivity reported a statistically significant and negative result for the year 2000, while reporting a positive but insignificant impact on labour productivity in 2003 for the FE model specification. Foreign Presence and Productivity Growth Girma et al. (2001) argue that the correct approach to measuring productivity spillovers in domestic firms is to focus on productivity changes in domestic firms rather than productivity levels. Domestic firms may exhibit higher levels of productivity in sectors with a large foreign presence but not higher levels of productivity growth. Taking this into account, the following labour productivity growth model is estimated for the indigenous software industry: DLP 5 b0 1 b1RD 1 b2FS 1 b3LQ 1 b4FP 1 Tt 1 Uit
(7.3)
where DLP is the annual change in labour productivity of indigenous firms between 2000 and 2003. RD, FS, LQ and Tt are measured as before and are estimated for the years 2000, 2001 and 2002. FP is the foreign
MNE subsidiaries, productivity spillovers and SMEs
Table 7.2
147
Foreign presence and labour productivity growth in the Irish indigenous software industry
Dependent variable labour productivity growth Independent variables RD FS LQ FP Year dummy 2000 Year dummy 2001 Year dummy 2002 Diagnostics R-squared Adjusted R-squared HN (RE v FE)a Condition index DW test (p-value) LM test (p-value) No. of obs N (firms)
Fixed effects (FE) coefficients
Random effects (RE) coefficients
0.010718 (1.27419) 0.000185241 (0.031812) 0.035298*** (7.129) 1.29834 (0.627812) −0.017396** (−3.26) −0.08612*** (−8.17) −0.05192*** (−9.26)
0.0067013 (1.45) −0.0013809 (−0.191057) 0.00963245*** (4.04985) 0.9417153 (0.497386) −0.00417 (−0.963415) −0.2183422*** −9.18) −0.49898*** (−6.87)
0.460475 0.167458 Fail to reject 5.32 1.97389 (0.098134) 5.87606 (0.361) 180 60
0.281652 0.231643 Reject 3.24 1.64111 (0.7431) 3.351 (0.3862) 180 60
Notes: ** significant at the 5 per cent level; *** significant at the 1 per cent level. Figures in parentheses are t-statistics unless otherwise stated. a HN: Hausman statistic for testing the random effects model against the fixed effects model; CHISQ (7) = 37.761, p-value= (0.000). Rejection of the random effects model. RD is research and development expenditure, FS is foreign sales, FP is foreign presence.
presence variable and is also measured as before (for the period 1997 to 2000). The results of the estimation are presented in Table 7.2. The Hausman test statistic (HN) indicates that the fixed effects model is the appropriate
148
SMEs in a globalised world
model in this instance, although the results are largely similar for both models in relation to statistical significance of the variables. The FP variable has a statistically insignificant effect on the growth in labour productivity. Foreign presence does not influence the growth in productivity of local software firms in the indigenous software industry. This finding suggests the absence of any productivity spillover effect for domestic firms from high levels of employment in the foreign software sector. The results show a positive and statistical significant relationship at the 1 per cent level for the labour quality variable. Labour quality is a key factor in driving labour productivity growth in the indigenous software industry. There is an insignificant relationship between labour productivity growth and the level of R&D expenditures. This is not in line with expectations but is a reasonable result given that it might take some time before high R&D expenditures translate into new innovations and then into productivity growth. The coefficients on the year dummies are statistically significant for the years 2001 and 2002, with a negative impact on firm labour productivity growth for both years at the one per level of significance. This confirms earlier results on the negative impact of these years for productivity growth. This estimated relationship is in contrast with the positive and statistically significant coefficient reported when the model was estimated in levels. This confirms to some extent the findings of Haddad and Harrison’s (1993) study which found that foreign presence contributed to levels of productivity in domestic firm but not to productivity growth rates.6 Absorptive Capacity, Labour Productivity Growth and Spillovers The finding that there is no relationship between labour productivity growth and foreign employment in the software industry requires further analysis. This section tests whether there is a difference across domestic firms in their ability to absorb spillovers and whether these differences affect the relationship between labour productivity growth in indigenous firms and foreign employment. The capability of host country firms to absorb foreign technology is important in determining the size of the spillover effect (Verspagen, 1993). This ability is positively associated with high levels of human capital, high levels of R&D expenditure, new product development and export intensity (Cohen and Levinthal, 1989; Mowery et al., 1996 and Kokko, 1996). Measuring absorptive capacity The data collected is useful in measuring the absorptive capacity of domestic firms. The approach here is to use four measures – export activity,
MNE subsidiaries, productivity spillovers and SMEs
149
R&D expenditures, innovation and human capital to more accurately capture the absorptive capacity of indigenous software firms. Indigenous firms in the sample are reclassified into firms with high, medium and low absorptive capacity using the measures listed above. Firms were allocated to the relevant subsample if they checked out positive in three out of the four absorptive capacity proxies. For the purposes of this research, high, medium and low absorptive capacity firms are defined as follows: 1.
2.
3.
High absorptive capacity. These are firms: ● with R&D expenditures as a percentage of sales greater than 15 per cent; ● that introduced a new product in the last three years; ● that have exports as a percentage of sales greater than 70 per cent; ● and have graduate employment more than 70 per cent of total employment. Ten firms were classified into this category. Medium absorptive capacity. These are firms: ● with R&D expenditures between 8 and 15 per cent of sales; ● that introduced a new product in the last three years; ● that have exports of between 30 and 70 per cent of sales; ● and have graduate employment between 40 and 70 per cent of total employment. Twenty-five firms were classified into this category. Low absorptive capacity. These are firms: ● with R&D expenditure below 8 per cent of sales; ● that introduced no new products in the last three years; ● that have exports below 30 per cent of sales; ● and have graduate employment below 40 per cent of total employment. Twenty-five firms classified into this category.
The cut-off points used in the classification of the sampled firms according to their absorptive capacity are not arbitrary. A study by Standard and Poors (2003) found that the average R&D expenditures in the global software industry is between 10 and 20 per cent of sales. This informs the choice of the high cut-off point at greater than 15 per cent of R&D expenditures for high absorptive capacity firms. The average exports as a percentage of sales in the indigenous software industry is 62 per cent according to Crone (2002). Companies which employ a higher percentage of graduates in total employment than the norm generally are more likely to be involved in product development and also be among the
150
SMEs in a globalised world
first to adopt leading-edge technology. Some surveys have put graduate percentages in software firms at between 70 and 90 per cent of total employment. The productivity growth equations for the three subsamples (low, medium and high absorptive capacity) were reestimated. The results of the estimations are presented in Table 7.3. It is only firms that are characterized by medium levels of absorptive capacity that benefit from high levels of foreign employment in the industry. Firms characterized by high and low absorptive capacity show insignificant coefficients on the foreign presence variable and hence do not benefit from a spillover effect from the level of foreign employment in the software sector. For medium absorptive capacity firms both the random effects and fixed effects models agree on the significance of the foreign presence variable but vary in relation to the magnitude of significance. The coefficient on the foreign presence variable is positive and significant at the 1 per cent level for the fixed effects model and the 5 per cent level in the random effects model. The Hausman test (HN) favours the fixed effects model for medium absorptive capacity firms, which supports a higher level of significance for the foreign presence variable. This result illustrates that the level of foreign employment influences positively productivity growth in these firms. Results also show that the level of labour quality is significant for productivity growth across all firms regardless of absorptive capacity. The employment of a skilled workforce improves annual changes in productivity. There is some difference in relation to the magnitude of the significance across the three subsamples. Firms with low absorptive capacity show a 1 per cent level of significance for this variable in both the random effects and fixed effects models (although the fixed effects model is favoured). Firms with medium absorptive capacity record a 1 per cent level of significance on the labour quality variable in the random effects model but only a 5 per cent level for the fixed effects model. However, the Hausman test indicated that the fixed effects model is best. For the high absorptive capacity firms, the labour quality variable is significant at the 5 per cent level for both models. Overall, the results accord well with the empirical literature. High levels of foreign employment in the software sector between 1997 and 2000 had a positive and significant impact on productivity growth for medium absorptive capacity firms. No such result was found for firms with low or high absorptive capacity. A key argument in the literature is whether large technology gaps (or low absorptive capacity) between foreign and domestic firms impeded or promoted the scope for local learning and hence the magnitude of spillovers to local firms. The findings support Cantwell’s (1993) argument that where the technology gap is very significant domestic
151
Year dummy 2002
Year dummy 2001
Year dummy 2000
ΔFP
ΔLQ
ΔFS
ΔRD
0.0028518 (0.244419) −0.0098039 (−0.37900) 0.02099*** (3.67192) −5.3975 (−1.4519) −0.083453 (−0.24519) −0.19588** (−3.13) −0.000678 (−1.6747)
(1) RE
(3) RE 0.00831501 (0.338159) 0.0061227 (0.980930) 0.026235*** (4.59866) 8.09144*** 6.20201 −0.01756 (−1.67660) −0.082544** (−3.55672) −0.041921** (−2.2243)
(4) FE
Absorptive capacity medium
0.00381295 0.012430 (0.279297) (1.19709) −0.00763316 0.00111781 (−0.401372) (0.691632) 0.018206*** 0.00827739** (5.62634) (2.95828) −2.25061 5.58519** (−0.545205) (2.94697) −0.060721 −0.0351798 (−1.44107) (−0.17634) −0.055520** −0.0091590** (−4.01355) (−3.84532) −0.038821** −0.0597721*** (−2.99234) (−6.3232884)
(2) FE
Absorptive capacity low (5) RE
0.00810423 (1.30915) 0.00425727 (1.07864) 0.010704** (3.18371) 0.597452 (0.292917) −0.0387887 (−1.23222) −0.0911113*** (−6.92544) −0.0042322** (−1.99)
(6) FE
Absorptive capacity high
0.00526493 (0.1967280 0.00017853 (0.110982) 0.0069731** (2.99710) 0.770507 (0.312565) −0.0034337 (−0.78978) −0.184838** (−2.9999) −0.095755*** (−5.17407)
Labour productivity growth for low, medium and high absorptive capacity firms
Dependent variable labour productivity growth
Table 7.3
152
(continued)
(1) RE 0.40125401 0.3550188 Reject RE CHISQ(7) = 15.8333 (p − 0.0033) 11.88 2.003 (0.337) 3.245 (0.61398) 75 25
(3) RE 0.494777 0.184754 Fail to reject CHISQ(7) = 15.8333 (p − 0.0033) 6.49 1.99074 (0.18) 3.33065 (0.18452) 75 25
(4) FE
Absorptive capacity medium
0.340524 0.219008 Fail to reject CHISQ (7) = 3.8084 (p = 0.4326) 10.12 1.57862 (0.329) 7.89862 (0.374) 30 10
(5) RE
0.431201 0.284281 Rejects CHISQ (7) = 3.8084 (p = 0.4326) 7.33 1.87843 (0.139) 2.33438 (0.127) 30 10
(6) FE
Absorptive capacity high
Notes: ** significant at the 5 per cent level; *** significant at the 1 per cent level. RD is research and development expenditure, FS is foreign sales, FP is foreign presence.
0.569311 0.307153 Fail to reject CHISQ (7) = 19.905 (p = 0.00005) 8.13 1.89 (0.173) 3.1004 (0.2160) 75 25
(2) FE
Absorptive capacity low
Diagnostics R-squared 0.340769 Adjusted R-squared 0.188187 HN (RE v FE) Reject RE CHISQ (7) = 19.905 (P = 0.00005) Condition index 9.54 Durbin Watson 1.45 (p-value) (0.9714) LM test 2.212 (p-value) (0.2666) No. of obs 75 N (firms) 25
Dependent variable labour productivity growth
Table 7.3
MNE subsidiaries, productivity spillovers and SMEs
153
firms will not benefit from foreign presence because a wide technology gap impairs indigenous firms ability to catch up. Similarly Blömstrom and Kokko (2001) and Verspagen (1993) find that very large differences in the technology gap between local and foreign firms inhibit spillovers. The finding of no productivity spillovers to domestic firms from high levels of employment in the foreign software sector on labour productivity growth in high absorptive capacity firms again supports Cantwell’s hypothesis. Cantwell finds that there must be some disparity in the level of technology between foreign and indigenous firms for productivity spillovers to occur. In addition, Blömstrom et al. (2000) show that in industries where the gap is negligible, or even where domestic firms are more advanced than MNE subsidiaries, one would not expect spillovers to occur.
WHAT ACCOUNTS FOR THE SPILLOVER EFFECT? There are a number of reasons why indigenous firms may be benefiting from productivity spillovers as a result of the presence of MNE subsidiaries in this sector. These include the presence of knowledge intensive export-oriented MNEs, absorptive capacity and product focus of indigenous firms, agglomeration effects and entrepreneurship. Knowledge Intensive Export-Oriented MNE Sector The literature on productivity spillovers suggest that knowledge intensive firms provide more opportunity for domestic firms to appropriate MNE technology through spillovers (Blomström et al., 2000; Kokko, 1996). This facilitates greater potential access for domestic firms to a supply of possible appropriable MNE technology (see Harris and Robinson, 2002). Almost all the top global software firms have a subsidiary in Ireland. It is likely that these subsidiaries bring with them part of the firm specific advantage of the parent company. Hence, there is potentially a large stock of foreign technology that could possibly spillover to domestic software firms. As already discussed above, the MNE software sector in Ireland is almost exclusively export oriented. Export-oriented MNE subsidiaries may result in larger productivity spillovers to local firms because foreign subsidiaries are not in direct competition with each other to control the host country market (Görg and Strobl, 2002). This provides scope for unlimited amounts of foreign capital to enter the host country bringing with them a diverse range of technologies that, if accessed by domestic firms, could potentially raise their productivity.
154
SMEs in a globalised world
Absorptive Capacity, Technological Capability and Product Focus There is no obvious benefit to indigenous firms from the presence of high technology foreign firms if they cannot understand or assimilate the technology of the subsidiary. The ability of local firms to assimilate MNE technology (absorptive capacity) and benefit from positive productivity spillovers can be measured by investment in human capital and R&D (Blomström et al., 2000; Liu, 2000). Using these measures as a yardstick, domestic software firms can be categorized as having a high absorptive capacity. Over 75 per cent of total employees in domestic firms are third level graduates (Enterprise Ireland, 2004) while expenditure on R&D averaged 8.4 per cent of revenue over the 2000–03 period (Buckley, 2005). High R&D expenditures in the indigenous sector are reflected in the product orientation of firms with over 44 per cent of firms involved in software development (Arora et al., 2001). The technological capability and absorptive capacity of domestic firms puts them in a unique position to understand and absorb MNE technology should it spill over. Clustering and Agglomeration Economies Agglomeration economies and the development of high technology clusters are also related to the finding of positive productivity spillovers (Barry et al., 2004). There is significant evidence of such agglomeration and clustering in the Irish software industry. For example, more than 70 per cent of MNE subsidiaries and 75 per cent of domestic firms are concentrated in or around the greater Dublin area (Crone, 2002). Such a concentration of indigenous and overseas software firms in close proximity to each other is likely to facilitate faster knowledge dissemination through, for example, the movement of workers between firms. Entrepreneurship There is evidence that foreign software companies are drivers of new firm start-ups in the indigenous sector. This is not surprising given the high technology nature of the Irish software industry. In Chapter 8 of this volume, Barbosa and Eiriz present empirical evidence to support the view that MNE presence has a positive effect on entry for local entrepreneurs in high technology industries in the Portuguese manufacturing and service sectors. Görg and Strobl (2002) empirically estimate a model describing the entry of indigenous firms in the Irish manufacturing sector. They find that foreign multinationals exert a positive effect on the entry of indigenous firms. De Backer and Sleuwaegen (2003) analysed firm entry and
MNE subsidiaries, productivity spillovers and SMEs
155
exit patterns across the Belgian manufacturing industry and found that the positive effects of FDI on domestic entrepreneurs in the long run as a result of learning, demonstration and linkage effects between foreign and domestic firms. This is supported by the findings of Giarratana et al. (2003) in their observation that the majority of domestic firms were set up after the entry of overseas companies. Fifty-seven per cent of all MNEs now located in Ireland accessed the Irish market prior to 1990. It was not until after 1990 that the rate of new local firm creation increased significantly and this upsurge in new domestic software firms followed entry by foreign software subsidiaries. The majority of companies were set up after 1993 with 59 per cent of all firms surveyed formed in the 1994–2000 period after entry by MNE subsidiaries in this sector. This is a possible indication of new local firm innovation as a result of a heavy foreign presence in the software sector. A survey of indigenous software firms by Hot Origin (2001) also finds that the majority of domestic firm start-ups in the Irish software industry came after the entry of overseas companies. Connected to this, is the finding that over 74 per cent of CEOs and or owners of ‘new’ domestic firms had former work experience in a MNE (Buckley, 2005). The movement of workers between foreign and domestic firms could well benefit domestic firms if managerial and technical expertise acquired in the multinational can be applied to raise the productivity levels of new local firms. Government Intervention Walz (1997) argues that government intervention can raise the potential for positive productivity spillovers from foreign presence. The Irish software industry has benefited significantly from interventions by the state. Such policies included an industrial strategy, which encouraged high technology MNE subsidiaries to locate in Ireland. The Irish ‘package’ of incentives include a low rate of corporate tax, investment in telecommunications, skill upgrading, capital and R&D grants and the creation of industry and trade associations. All of these were aimed at creating a competitive advantage in local software production. These interventions add to the potential for productivity spillovers through building-up absorptive capacity in the sector. Although not focusing on the software industry per se, Lenihan et al. in Chapter 10 in this volume on the dynamics of the SME sector in Ireland, do however, begin to question the true costs of the various industrial policy interventions by Irish policymakers. This is an issue which is clearly beyond the scope of the current chapter but nonetheless worth mentioning in the current context.
156
SMEs in a globalised world
CONCLUSIONS This chapter tested whether foreign MNE subsidiaries in the Irish software sector benefited indigenous software firms through improvements in productivity. The Irish software industry is an interesting choice given that it is composed of large foreign firms alongside an indigenous sector that is largely dominated by SMEs. Previous research on productivity spillovers focuses almost exclusively on assessing the impact of foreign MNE subsidiaries on large domestic manufacturing firms and uses highly aggregated data. As such, not much is known about the impact of large foreign subsidiaries on indigenous SMEs especially with respect to productivity. This research remedies the deficiency and adds to the current body of knowledge. A labour productivity model was estimated using panel data techniques and data for the model was collected from a detailed survey of local software firms. The model was initially estimated in levels and found strong evidence in favour of productivity spillovers. The model was extended to reflect some of the key issues regarding the magnitude of spillovers. The first extension involved re-estimating the model to take account of the effect of the impact of high levels of foreign presence on labour productivity growth. However, the coefficient on the foreign presence variable was statistically insignificant indicating the absence of a productivity spillover effect. A second extension of the model takes account of the key debate in the literature with respect to absorptive capacity and the technology gap. Medium absorptive capacity firms in the indigenous software industry benefit most from high levels of employment in the foreign software sector. The results indicate, in support of Cantwell, that some difference in technology levels is required for domestic firms to benefit from the presence of foreign firms through a productivity effect. Another interesting finding as a result of the research was the importance of both skills and qualifications to the level and growth in productivity of domestic firms, absorptive capacity aside. This is especially true in the case of the software industry because of the pace at which skills in this sector become outdated as new innovations, products and processes are continually launched on the market. An important question to address is what accounts for the productivity spillover effect to domestic firms? The arguments discussed earlier are relevant here. The significant presence of a knowledge intensive industry that is almost wholly export-oriented is very relevant. These firms are likely to use leading edge technologies that domestic firms may not have been aware of, or have access to, prior to foreign entry. However, such technology is of little use to domestic firms if they cannot understand and apply it to improve their own productivity. The significant technological
MNE subsidiaries, productivity spillovers and SMEs
157
capability and the very high skill level characteristic of indigenous firms were undoubtedly important in this respect. The government too played a strong role in attracting these firms as well as building-up the absorptive capacity of domestic firms and increasing their chances of absorbing any spillovers from the overseas software sector. The low capital intensity of the general software sector along with low barriers to entry provided an opportunity for enterprising individuals to use their MNE work experience to best advantage by setting up their own firm. The experience of the Irish indigenous software industry provides ample evidence that indigenous SMEs in a peripheral region in the EU can successfully survive and flourish alongside a strong foreign MNE sector as a result of learning and demonstration effects between foreign and domestic firms. However, these benefits from MNE presence are not universal or automatic. On the one hand, SMEs with low technological capability may be significantly disadvantaged by foreign presence as they simply cannot compete with the superior technology of foreign firms. On the other hand, SMEs with a technological capability on a par with their foreign counterparts have limited opportunities to benefit from productivity spillovers as the technology gap between the two is too narrow. This chapter allows us to draw important lessons regarding the extent to which SMEs benefit from productivity spillovers due to the presence of large foreign-owned MNE subsidiaries It also highlights the conditions under which these benefits may be maximized and as such may provide important guidelines for governments wishing to pursue an industrial policy based on attracting multinationals. Irish industrial policy has been extremely successful in generating growth in what once was a weak peripheral economy on the edge of Europe. Moreover, the Irish model has not endangered the growth and development of SMEs in the high technology sector, in fact, these have grown substantially, especially since the mid-1990s. In this regard, other new member states may extract valuable lessons from the Irish experience which could well increase their pace of economic development.
NOTES 1. According to a National Informatics Directorate (2006) many of these firms are very small firms, employing less than five people and these can be more difficult to track. 2. Ideally, it would be interesting to include these firms as part of the sample, however, no CEO or co-owner or other high-level managers were interested in participating in this study. 3. A few studies do focus on other sectors, apart from manufacturing. For example, Chung et al. (1998) examine the influence of Japanese FDI on the US automotive component industry.
158
SMEs in a globalised world
4. Diagnostic testing indicates that the models are free from heteroskedasticity, autocorrelation and multicollinearity and are reported in detail in Table 7.1. 5. Although the fixed effects model is the preferred specification all models are in agreement with respect to the statistical significance of the variables. 6. However, Haddad and Harrison (1993) found no positive effect from foreign presence on labour productivity growth in local firms.
REFERENCES Arora, A., A. Gambardella and S. Torrisi (2001), ‘In the footsteps of Silicon Valley? Indian and Irish software in the international division of labour’, Stanford Institute for Economic Policy Research discussion paper no. 00-41, Stanford University. Baltagi, B. (2001), Econometric Analysis of Panel Data, 2nd edn, New York: Wiley. Barry F., H. Görg and E. Strobl (2004), ‘Multinationals and training: some evidence from Irish manufacturing industries’, Scottish Journal of Political Economy, 51 (1), 49–61. Blomström, M. and A. Kokko (2001), ‘Foreign direct investment and spillovers of technology’, International Journal of Technology Management, 22 (5/6), 435–54. Blomström, M., A. Kokko and S. Globerman (2000), ‘The determinants of host country spillovers from foreign direct investment’, in N. Pain (ed.), Inward Investment, Technological Change and Growth, Basingstoke: Palgrave Macmillan, pp. 34–65. Buckley, R. (2005), ‘Multinational enterprises, productivity spillovers and the growth and development of the Irish software industry’, unpublished PhD thesis, University of Limerick. Cantwell, J. (1993), ‘Technological competence and evolving patterns of international production’, in H. Cox, J. Clegg and G. Letto-Giles (eds), The Growth of Global Business, London: Routledge. Caves, R.E. (1996), Multinational Enterprise and Economic Analysis, 2nd edn, New York: Cambridge University Press. Chung, W., W. Mitchell and B. Yeung (2003), ‘Foreign direct investment and host country productivity: the American automotive component industry in the 1980s’, Journal of International Business, 34, 199–218. Coe, N. (1997), ‘US transnationals and the Irish software industry: assessing the nature, quality and stability of a new wave of foreign direct investment’, European Urban and Regional Studies, 4 (3), 211–30. Cohen, W.M. and D.A Levinthal (1989), ‘Innovation and learning: the two faces of R&D’, Economic Journal, 9 (397), 569–96. Crone, M. (2002), ‘A profile of the Irish software industry’, consultation draft published on line, accessed 26 February 2007 at: www.qub.ac.uk/nierc. De Backer, K.D. and Leo Sleuwaegan (2003), ‘Does foreign direct investment crowd out domestic entrepreneurship?, Review of Industrial Organisation, 22 (1), 67–84. Enterprise Ireland (2004), Software Industry Statistics for 1991–2003, accessed 16 March 2007 at www.nsd.ie/ht,/ssii/stat.htm. Eurostat (2008), Europe in Figures: Eurostat Yearbook, Luxembourg: European Commission.
MNE subsidiaries, productivity spillovers and SMEs
159
Giarratana, M., A. Pagano and S. Torrisi (2003), ‘Links between multinational firms and domestic firms: a comparison of software in India, Ireland and Israel’, paper prepared for the conference Reinventing Regions in a Global Economy, 12–15 April, Pisa, Italy. Girma, S., D. Greenaway and R. Kneller (2004), ‘Does exporting lead to better performance: a microeconometric analysis of matched firms’, Review of International Economics, 12, 855–66. Girma, S., D. Greenway and K. Wakelin (2001), ‘Who benefits from foreign direct investment in the UK?’, Scottish Journal of Political Economy, 48 (2), 119–33. Görg, H. and D. Greenaway (2004), ‘Much ado about nothing? Do domestic firms really benefit from foreign direct investment?’, World Bank Research Observer, 19 (2), 171–97. Görg, H. and E. Strobl (2002), ‘Multinational companies and indigenous development: an empirical analysis’, European Economic Review, 46 (7), 1305–22. Görg, H., A. Hijzen and B. Murakozy (2006), ‘The productivity spillover effect of foreign owned firms: firm level evidence for Hungary’, University of Nottingham research paper no. 2006/08. Haddad, M. and A. Harrison (1993), ‘Are there positive spillovers from FDI: evidence from Morocco’, Journal of Development Economics, 42 (1), 51–74. Hale, G. and C. Long, (2007), ‘Are there productivity spillovers from foreign direct investment in China’, Federal Reserve Bank of San Francisco working paper series no. 2006-13. Harris, R. and C. Robinson (2002), ‘Spillovers from foreign ownership in the United Kingdom: estimated for UK manufacturing using the ARD’, University of Durham working paper. Hot Origin (2001), ‘Ireland’s emerging software cluster: a hothouse of future stars’, report overview accessed 12 June 2007 at: www.hotorign.com. Irish Software Association (ISA) (2007), Annual Review 2006–07, Dublin: ISA. Javorcik, B. and S. Smarzynska (2004), ‘Does foreign direct investment increase the productivity of domestic firms? In search of spillovers through backward linkages’, American Economic Review, 94 (3), 605–27. Kokko, A. (1996), ‘Productivity spillovers from competition between local firms and foreign affiliates’, Journal of International Development, 8 (4), 517–30. Lipsey, Robert E. and F. Sjöholm (2005), ‘The impact of inward FDI on host countries: why such different answers?’, in T.H. Moran, E.M. Graham and M. Blomström (eds), Does Foreign Direct Investment Promote Development?, Washington, DC: Institute for International Economics. Liu, Z. (2000), ‘Foreign direct investment and technology spillovers: some evidence from China’, presentation to the Deng’s Nan Xun Legacy and China’s Development Conference, 11–13 April 2000, Singapore. Mowery, D.C., J.E. Oxley and A. Silverman (1996), ‘Strategic alliances and inter firm knowledge transfer’, Strategic Management Journal, 17 (Winter special issue), 77–91. National Informatics Directorate (NID) (2006), ‘Profile of the Irish owned software industry’, accessed 22 March 2007 at www.nsd.ie/htm/its/profile.htm. O’Gorman, C., E. O’Malley and J. Mooney (1997), ‘The Irish indigenous software industry – an application of Porter’s cluster analysis’, National Economic and Social Council research series paper, no.3, Dublin. O’Riain, S. (2000), ‘the flexible development state: globalisation, information technology and the Celtic Tiger’, Politics and Society, 28 (2), 157–93.
160
SMEs in a globalised world
Scott-Kennel, J. and P. Enderwick (2006), ‘FDI and interfirm linkages: exploring the black box of the investment development path’ Transnational Corporations, 14 (1), 1–33. Standard and Poors (2003), Industry Survey Computers: Software, Dublin: McGraw-Hill. Verspagen, B. (1993), Uneven Growth Between Interdependent Economies, Aldershot: Avebury. Walz, U. (1997), ‘Innovation, foreign direct investment and growth’, Economica, 64, 63–79. Wei, Y. and X. Liu (2006), ‘Productivity spillovers from R&D, exports and FDI in Chinese manufacturing sector’, Journal of International Business Studies, 37 (4), 544–57.
8.
Entrepreneurship and inward foreign direct investment in Portugal1 Natália Barbosa and Vasco Eiriz
INTRODUCTION Over the past decades governments across the world, both in developing and developed countries have identified foreign direct investment (FDI) as a key factor of their national development strategies, designing policies and incentives that increasingly facilitate the location of multinational enterprises (MNEs) in their territories. The rationale for this is based on the argument that MNEs can be influential to the host economy by assisting and promoting local entrepreneurial activity and industrial restructuring. The entry or presence of MNEs is thought to provoke a positive effect on local entry and post-entry performance in terms of survival and growth of new enterprises by increasing demand for intermediate inputs and, hence, by yielding pecuniary spillovers. Rivera-Batiz and Rivera-Batiz (1990), Rodríguez-Clare (1996), Markusen and Venables (1999), and Barrios et al. (2005) provided theoretical treatments of the main mechanisms by which MNEs may uphold the development of local enterprises and industries. This chapter analyses empirically whether MNEs operating in Portugal have a positive and important effect on local entrepreneurial activity. The relationship between MNEs and the creation of new local enterprises is investigated in context of both manufacturing and services industries and for different groups of industries (classified according to Pavitt, 1984) and small- and medium-sized enterprises (SMEs). The Portuguese economy provides an interesting context for such an analysis given that over the past decades, after Portugal’s entry into the European Union (EU) in 1986, the expected positive effects of FDI on local development were taken as to justify generous financial and fiscal incentive packages for attracting MNEs. Moreover, it is a small, open and developed economy, which is mainly populated by SMEs. It should be noted, however, that there is not a significant difference between the Portuguese SME sector and the SME 161
162
SMEs in a globalised world
sector of the average OECD country (Cabral, 2007). In the context of the EU, Portugal is a peripheral economy attracting MNEs that have been mostly searching locations with relatively low production costs (Barbosa et al., 2004). The motivations for FDI in Portugal may be distinct from those in other European countries, which may thus affect the relationship between FDI and the scale and nature of opportunities to new would beentrepreneurs (Driffield and Love, 2006). The concern about the motivations of FDI in Portugal is of substantial interest as previous studies revealed that MNEs operating in Portugal appear to behave differently from MNEs in other European countries. In a study of the entry process of MNEs in the Portuguese manufacturing industry, Barbosa et al. (2004) found that MNEs in Portugal are able to ignore prominent local market features (for example, averaged profitability, product differentiation and industrial concentration), desiring low wages and industries where other MNEs are clustered. Mata and Portugal (2004) have also showed that new Portuguese enterprises are typically smaller than new MNEs and, more importantly, they also have different sectoral entry patterns. Specifically, MNEs enter into industries where the protection offered by entry barriers acts as a selection mechanism, precluding the entry of local entrepreneurs. These results suggest that the Portuguese experience reflects an entry process of MNEs that use the country as a platform for their export-oriented activities, revealing weak linkages to the local market. Another study shows that most of the MNEs in Portugal opt for full foreign-ownership, lessening the possible linkages with local enterprises (Barbosa and Louri, 2002). This preference in terms of ownership suggests that MNEs in Portugal are concerned with potential spillover costs. They also appear not to be interested in accessing local market knowledge through partnerships with local enterprises. Another puzzling result is also found with respect to corporate performance. MNEs in Portugal do not perform better than local-based enterprises (Barbosa and Louri, 2005), while in other countries MNEs appear to attain a relative superior performance. Using enterprise level panel data for manufacturing and services industries in Portugal and applying parametric and non-parametric econometric techniques, we only found support for a u-shaped relationship between MNEs and the creation of local enterprises (as posited by theoretical models) in the case of high-tech industries. For some other cases, the estimates suggest a linear relationship but, more importantly, a negative impact of inward FDI on the net creation of local enterprises. In particular, the results indicate that the net creation of enterprises into the Portuguese SME sector is negatively affected by MNEs. The competition
Entrepreneurship and inward FDI in Portugal
163
challenge associated with MNEs in Portugal seems to constrain the evolution of the Portuguese SME sector in terms of size and growth. Nonetheless, this result can be interpreted as a positive impact of MNEs. In fact, it may induce an industrial restructuring process in which efficient enterprises are forced to exit and are replaced by more efficient enterprises. If this is the case, Portuguese enterprises appear to respond positively to the challenges of globalization by interpreting unfavourable productivity signals as a need to exit and reallocate resources. On the other hand, enterprises receiving favourable information about their efficiency may use the presence of MNEs as a channel to access new markets and competitive competences. The chapter concludes by discussing the policy implications emerging from the results, which may assist SMEs to respond better to the challenges of globalization and industrial restructuring.
THEORETICAL FRAMEWORK Theoretical treatments of the effects of inward FDI on local industrial development (see, for example, Markusen and Venables, 1999; Barrios et al., 2005) show that MNEs have two main effects on host economies, apart from the positive macroeconomics effect of a larger capital endowment in the host economy. The effect of MNEs on host economies manifest themselves in two ways: (a) a demand effect and (b) a competition effect. Entry of MNEs raises the demand for local intermediate goods. This impact depends, however, on the input–output linkages MNEs generate compared to the linkages local enterprises would themselves generate. When MNEs replicate local enterprises linkages this leads to a quantity effect on the demand for intermediate goods. On the other hand, when MNEs have higher linkages coefficients than local enterprises, this tends to increase the variety of intermediate inputs produced locally, leading to a beneficial effect on the local economy as a whole. In both cases, MNEs are important in assisting the development and creation of local enterprises, but in the latter case MNEs can also induce changes in the local industrial structure. In the second effect, MNEs are competing with local producers on their product market as well as on the factor market. The product competition pressure on local enterprises tends to lower the price index of the final product, thereby forcing local inefficient enterprises to exit the market. As a consequence, the entry of MNEs lowers the demand from local enterprises for intermediate goods. A positive combined effect occurs when the demand effect dominates the competition effect. The potential gain for the host economy in terms of local entrepreneurial activity then lies in the
164
SMEs in a globalised world
relative strength of the positive externality associated to the demand effect and the intensity of competition pressure in the product and the factor markets. Moreover, this positive effect requires that the MNEs’ technological advantage in producing the final good is not too large vis-à-vis its local competitors (Lin and Saggi, 2005). Based on numerical simulations of the equilibrium values for the number of active enterprises (local and MNEs), Barrios et al. (2005) showed that the main economic forces at work lead to a u-shaped relationship for the evolution of the number of local enterprises as a function of the presence of MNEs. In an initial phase, the competition effect dominates the demand effect and a reduction in the number of local enterprises may be observed. However, as the number of MNEs increases, the competition effect is gradually outweighed by positive externalities associated with the demand effect. This suggests that there is a minimum number of MNEs that should be operating in a host economy in order to provoke positive externalities in terms of promoting local development. Moreover, the sectoral distribution of this minimum number of MNEs should be strongly connected with the input–output linkages MNEs generate in the host economy. The motivations of FDI could also determine the effect of MNE entry on the host economy. In fact, MNEs are not equal in their skills, knowledge base, the approaches they use to protect knowledge, their strategic orientation in the entry markets and, consequently, in the opportunities they offer to local entrepreneurs (Almeida and Phene, 2004; Adler and Hashai, 2007). For instance, export-oriented FDI, in which the host economy serves as a production platform, is likely to enlarge the positive effect of MNEs on local entrepreneurship. In this case, the competition effect is mitigated as MNEs only exert competition pressure on the factor market. They may in fact allow local enterprises in the intermediate goods sector to expand their production via the exports of MNEs (Barrios et al., 2005) or by supplying other affiliates belonging to the MNE production network. On the other hand, market-seeking FDI is likely to increase the level of competition in the host country, reducing local enterprises’ sales and driving some local enterprises to exit the market. Although this may promote local efficiency levels and innovation, market-seeking FDI tends to accentuate the negative effect regarding the local net entry rate. The majority of previous empirical studies that have investigated the effects of inward FDI on local enterprises have mainly focused on local productivity gains and growth as a result of MNEs entry. Nonetheless, robust empirical evidence on productivity spillovers from inward FDI is hard to find. Görg and Greenaway (2004) identified 35 papers on productivity spillovers from multinationals to local enterprises for developing countries, developed countries and transition countries. Most of these
Entrepreneurship and inward FDI in Portugal
165
studies did not find evidence to support positive spillovers from MNEs to local enterprises. That is, they fail to show that local enterprises’ efficiency is an increasing function of MNEs’ presence in the host economy. In fact, some of the studies find either negative or statistically insignificant effects (see, for example, Aitken and Harrison (1999) for Venezuelan industry; Kathuria (2000) for Indian manufacturing industry; Girma et al. (2001) for the British industry; and Barrios and Strobl (2002) for the Spanish industry). In the case of Portuguese industry, Flores et al. (2007) found a positive relationship between local enterprises’ productivity and MNEs but only when the technological gap between them is not too small (otherwise there is little room for improvement) or too wide. However, the existence of a proper technological gap is a necessary but not sufficient condition. The industry-specific characteristics are a crucial influence for productivity spillovers to take place, leading to modest productivity effects of MNEs on the Portuguese economy. Another important effect of inward FDI on local enterprises can be associated with long-term economic restructuring. In fact, as we have previously discussed, the presence of MNEs on a host economy can lead to opportunities for the creation of new local enterprises. Despite the relevance of this effect, the evidence on the effects of MNEs’ entry on local entrepreneurship is quite scarce and has been generating contradictory results. De Backer and Sleuwaegen (2003) have found that an immediate impact of import competition and inward FDI may discourage entry and stimulate the exit of local entrepreneurs. Conversely, Cantwell and Piscitello (2002), McKeon et al. (2004) and Görg and Strobl (2005) have shown that MNEs’ entry encourages local enterprise creation. However, none of these studies looks at the nature and type of the local enterprises created. This issue is important because different types of enterprises may respond differently to the challenges of globalization, impacting differently on the local economic and technological development. In the next section of the chapter, we examine how local entrepreneurs respond to the opportunities created by the MNEs operating in Portugal. We look at the number and types of new local enterprises.
MODELLING THE LOCAL ENTREPRENEURSHIP– FDI RELATIONSHIP Empirical Specification Differences in local enterprises’ entry rate across industries can be related to differences in industry-specific characteristics, such as industry growth
166
SMEs in a globalised world
and competition, as well as differences in industry attractiveness to MNEs. Positioning this study in the tradition of entry models in empirical industrial organization (for example, Acs and Audretsch, 1989; Geroski, 1991; Mata, 1993), we specify that the local net entry rate in industry j at time t has additively separate components of the form: Ejt 5 dFDIjt 1 bh (FS_Ejt) 1 gzjt 1 aj 1 ft 1 ejt
(8.1)
where Ejt is the net entry rate defined as the number of local enterprises entries minus exits over the period t − 1 to t divided by the total number of enterprises at time t − 1 in industry j, zjt is a vector of time and industry varying covariates. The latter comprises industry size (lnSIZE) measured through the total number of employees, industry annual growth rate (IGROW), a minimum efficient scale (MES) measure defined by the average size of enterprises in an industry, the Herfindahl index (HHI)2 of industry concentration using enterprises’ employment as a good proxy for enterprises’ market share,3 and average age of enterprises in an industry (AGE). The choice of these industry-specific covariates is supported by previous studies on enterprise entry (for example, Mata, 1993; Geroski, 1995; Fotopoulos and Spence, 1999). aj are the unobserved industryspecific time-invariant effects which allow for heterogeneity in the means of the Ejt series across industries, while ft are the unobserved time varying effects, such as macroeconomic shocks, that impact on all industries. ejt are disturbance terms. The main covariates of interest are FDI and FS_E, both intending to capture the effect of MNEs operating in Portugal on the entry of new local enterprises. FDI is a dummy variable taking the value 1 if there is at least one MNE (that is, an enterprise where foreign investors own more that 10 per cent of the equity in the enterprise) in industry j at time t, while FS_E is computed as the share of employment in MNEs to total employment at industry level (that is employment in MNEs weighted by foreign equity participation of the enterprise divided by total employment in industry j at time t). The combination of these two covariates in the entry model allows us to test whether the impact of MNEs on local enterprises’ entry depends on the number and scale of previous MNEs. As the marginal impact of MNEs on local enterprises’ entry is likely to change over time (due to the changing market structure induced by the entry of new local and multinational enterprises) the inclusion of the cumulated number of MNEs in industry j at time t (FS_Ejt) allows us to disentangle whether the leading effect on entry is due to the presence of MNEs or to the intensity of such presence. Theory usually considers the number of MNEs as the variable of
Entrepreneurship and inward FDI in Portugal
167
interest. This variable has however important drawbacks associated with it as follows: (a) it does not take into account MNEs’ diversity in terms of size; (b) it neglects the empirical fact that MNEs are generally larger than local enterprises (for example, Barry and Bradley, 1997 for Ireland; and Barbosa and Louri, 2005 for Portugal and Greece); (c) it ignores the fact that the impact of MNEs on the host economy depends on the type of relationships they establish with local enterprises. Hence, in order to overcome these drawbacks we measure the importance of MNEs in an industry by the share of employment in MNEs, weighted by the percentage of enterprises’ equity owned by foreign investors. The relationship between the entry rate of local enterprises and MNEs’ presence at industry level in Equation 8.1, specified by the function h(.), can take any functional form. Bearing in mind the theoretical prediction about that relationship, we parametrically estimate Equation 8.1 assuming a u-shaped relationship. We also allow the data to determine the true functional form of that relationship by applying a semi-parametric estimator to the entry model. The advantage of this procedure is that no ex ante relationship is imposed to the data and, more importantly, the true relationship in the Portuguese case can emerge from the data and not be constrained by empirical evidence gathered in other countries. The Data Our enterprise level data source is the annual Quadros de Pessoal survey carried out since the early 1980s by the Portuguese Ministry of Labour and Social Security. We had access to the data from 1985 until 2000. The mandatory nature of this survey ensures that virtually the whole population of enterprises with wage earners in Portugal is included in the Quadros de Pessoal dataset. Data at enterprise level includes the percentage of enterprise’s equity owned by foreign investors, number of employees, characteristics of employees in terms of skills and academic qualifications, location, year of start-up, and the main economic activity performed by enterprises. For the purpose of this study we mainly use data on employment to compute our empirical variables. Data on the percentage of the enterprise’s equity owned by foreign investors form the basis for the identification of the presence and scale of MNEs at industry level. The aggregation at industry level is based on NACE five-digit classification of industries. Tables 8.1 and 8.2 present a summary of descriptive statistics of the empirical variables for the manufacturing and services industries, respectively. The descriptive statistics show that local enterprises’ entry rate is, on
168
SMEs in a globalised world
Table 8.1
Descriptive statistics of variables for the Portuguese manufacturing industry
Variable Local entry rate FDI FS_E lnSIZE MES HHI IGROW AGE
No. of obs
Mean
Std dev.
Min.
Max.
4122 4122 4122 4122 4122 4122 4122 4110
0.036 0.515 0.103 6.303 82.386 0.311 0.067 54.741
0.337 0.500 0.195 1.930 299.525 0.309 0.815 20.772
−1 0 0 1.609 5 0.001 −0.996 10.476
5 1 1 11.832 9624 1 21.200 104.882
Note: The explanatory variables are FDI – a dummy variable taking the value 1 if there is at least one multinational enterprise in an industry, FS-E – the share of employment in multinational enterprise to total employment at industry level, lnSIZE – industry size measured by the logarithm of the total number of employees, MES – a measure of minimum efficient scale at industry level, HHI – the Herfindahl index of industry concentration, IGROW – industry annual growth rate, AGE – average age of enterprise in an industry. Source:
Authors’ calculations based on the Quadros de Pessoal database (1985–2000).
Table 8.2
Descriptive statistics of variables for the Portuguese services industry
Variable Local entry rate FDI FS_E lnSIZE MES HHI IGROW AGE
No. of obs
Mean
Std dev.
Min.
2347 2347 2347 2347 2347 2347 2347 2337
0.072 0.590 0.063 6.564 50.690 0.173 0.104 49.938
0.341 0.492 0.116 2.006 236.468 0.263 0.585 13.700
−1 0 0 1.609 5 0.001 −0.933 4.146
Max. 4 1 0.871 10.980 3103.067 1 12.225 71.083
Note: The explanatory variables are FDI – a dummy variable taking the value 1 if there is at least one multinational enterprise in an industry, FS-E – the share of employment in multinational enterprise to total employment at industry level, lnSIZE – industry size measured by the logarithm of the total number of employees, MES – a measure of minimum efficient scale at industry level, HHI – the Herfindahl index of industry concentration, IGROW – industry annual growth rate, AGE – average age of enterprise in an industry. Source:
Authors’ calculations based on the Quadros de Pessoal database (1985–2000).
Entrepreneurship and inward FDI in Portugal
169
average, 3.6 per cent in the manufacturing industry and twice that in the services industry. The presence of MNEs at industry level is similar in both cases. Roughly 50 per cent of observed industries report the presence of at least one MNE. However, when the number of employees is taken into account in measuring the importance of MNEs, the manufacturing industry shows a greater presence of MNEs than the services industry. On the other hand, not surprisingly over the period 1986–2000, the average annual growth rate is greater in services than in manufacturing. Another interesting feature is that the competition intensity measured by the Herfindahl index (HHI) is greater in services than in the manufacturing industry. Regarding industry size, on average each observed services industry employs 709 workers, while each observed manufacturing industry employs 546 workers. However, the minimum efficient scale (MES) in the manufacturing industry is 64 per cent larger than in the services industry. This indicates that the population of enterprises in each industry is quite different. The manufacturing industries appear to comprise, on average, a small number of enterprises, but these are larger than those in the services industries. Parametric Approach In order to parametrically estimate the entry model we use the Generalized Method of Moments (GMM) methodology. The definition of the empirical variables suggests that both the dependent variable and covariates are correlated with the industry-specific effects aj. The advantage of using panel data is that consistent parameter estimates can be obtained in the presence of this kind of unobserved heterogeneity. To eliminate the industry-specific effects, a transformation such as first-differencing is required in order to obtain valid moment conditions. Therefore, we estimate the first-differencing version of Equation 8.1 DEjt21 5 dDFDIjt21 1 bh (DFS_Ejt21) 1 gDzjt21 1 ft 1 Dejt21
(8.2)
thus eliminating a potential source of omitted variable bias in estimation. Another important econometric issue is the potential endogeneity problem. Maintaining the assumption that the disturbances ejt are serially uncorrelated, the series of covariates may be endogenous in the sense that they are correlated with ejt and earlier shocks. That is, unobserved industry and time-varying shocks may impact on MNEs’ entry and continuing decisions as well as on industry characteristics. If this is the case, then their lagged values (for example, zjt-2, zjt-3 and longer lags) will be valid instruments in the first-differenced equations for periods t = 3, 4, . . ., T. That is,
170
SMEs in a globalised world
the endogenous variables in first differences are instrumented with suitable lags of their own levels. In this context, we apply the Arellano and Bond (1991) estimator. The key identifying assumption that there is no serial correlation in the disturbances is assessed by testing for no second-order serial correlation in the first-differenced residuals (that is, the Arellano–Bond test for AR(2) in first differences). In particular, negative first-order serial correlation is expected in the residuals if the disturbances ejt are serially uncorrelated. But higher-order autocorrelation is unexpected as it indicates that some lags of the dependent variable, which might be used as instruments, are in fact endogenous and thus considered bad instruments. Semi-parametric Approach The main advantage of a semi-parametric approach to examine the relationship between entrepreneurship and inward FDI is that it allows the data to identify the true functional form, accounting for the possible nonlinearity of h(FS_Ejt) and, simultaneously, for the effect of other covariates (zjt). In particular, we apply the Robinson’s (1988) semi-parametric Kernel regression estimator, allowing h(.) to be a smooth and continuous function of FS_E. This estimator requires that, in an initial step, the effects of the other covariates have to be removed from the relationship between FS_E and E. For that, we implement the (non-parametric) Nadaraya–Watson estimator modified from Salgado-Ugarte et al. (1994) and based on the equations provided by Härdle (1991) and Scott (1992) to estimate E(Δzjt−1/ FS_E) and E(ΔEjt−1/FS_E). The estimation of a Kernel regression requires the choice of the bandwidth (the smoothing parameter) and the Kernel function. We use the Gaussian Kernel (weight) function and the optimal bandwidth, which minimizes a cross-validation criterion. Note that Härdle et al. (2004) have argued that in estimating Kernel regressions the choice of the bandwidth is more important for the optimality of the estimator rather than the Kernel function. In the next step, we estimate the equation DEjt21 2 E^ (DEjt21 /FS_Ejt21) 5 q (Dzjt21 2 E^ (Dzjt21 /FS_Ejt)) 1 xjt21 (8.3) where xjt−1 is a random error term. Then, we use the estimates q^ , E^ (DEjt21 /FSEjt) and E^ (Dzjt21 /FS_Ejt) to compute h^ (FS_Ejt) 5 E^ (DEjt21 /FS_Ejt) 2 q^ E^ (Dzjt21 /FS_Ejt)
(8.4)
Entrepreneurship and inward FDI in Portugal
171
The semi-parametric Kernel regression estimates of h(FS_E) are then predicted values of entry rate from which the effects of z have been removed. To determine the estimated slope and shaping of the local entrepreneurship–FDI relationship we plot the predicted values of entry rate (purged from other covariates effects) and the intensity of MNEs presence at industry level (FS_E). Discussion of Results To assess whether inward FDI is influential to the Portuguese economy by promoting local entrepreneurial activity, we estimate Equation 8.1 by including the indicator of foreign presence (FDI), the intensity of foreign presence (FS_E) and its second-uncentred moment value in the set of covariates. These covariates will assess the impact of the first foreign investment on local entry rate and the cumulative impact of MNEs as the number and scale of foreign investments at industry level increases. This dynamic effect could be crucial to understand the impact of FDI on the development of Portuguese industry and to clarify and match previous empirical results on the effect of FDI on the Portuguese economy. The estimation results for all manufacturing and services industries are depicted in Table 8.3. First, we assess the validity of our specification through the Sargan test for over-identifying restrictions and the Arellano and Bond (1991) tests for first- and second-order serial correlation. For all regressions, the Sargan test validates the adequacy of instruments given that the over-identifying restrictions cannot be rejected. As expected, the Arellano and Bond test for first-order serial autocorrelation is found to be negative and significant at the 1 per cent level, while the test for second-order autocorrelation is rejected. This confirms the consistency of the GMM estimator and hence the appropriate econometric specification of the entry model. For the manufacturing industry, the impact of the first MNE is positive and statistically significant at the 10 per cent level. This suggests that the linkage effect through intermediate goods demand dominates when the first MNE decides to be an active enterprise in Portugal. However, as the number of MNEs increases, the competition effect appears to dominate any positive externality associated to a demand increase on the intermediate goods industry. The positive, though weakly significant coefficients associated with FDI suggests that, initially, the entry of MNEs can kick-start local entrepreneurship, but its positive effect is quickly erased as the number and scale of MNEs increase. By including a quadratic term of FS_E and excluding the first effect associated with MNEs entry, this renders the linear term of FS_E positive
172
Table 8.3
SMEs in a globalised world
GMM estimation of annual local entry rate in the period 1986–2000 Manufacturing
FDI FS_E FS_E2 lnSIZE MES HHI IGROW AGE Observations m1 m2 Sargan test (p-value)
(1) 0.176* (0.097) −0.969*** (0.313) –
(2) –
Services (1) 0.126 (0.184) −2.295** (0.966) –
(2) –
0.295 −1.556 (0.690) (1.508) −1.544* −1.961 (0.834) (2.125) 0.199* 0.202* 0.096 0.016 (0.113) (0.109) (0.233) (0.242) −0.000 −0.000 −0.000 0.000 (0.000) (0.000) (0.000) (0.000) −1.342*** −1.261*** −1.850*** −1.927*** (0.459) (0.428) (0.526) (0.521) 0.170** 0.172** 0.429*** 0.458*** (0.072) (0.072) (0.117) (0.117) 0.017** 0.018** 0.011 0.013 (0.007) (0.007) (0.012) (0.012) 3116 3116 1791 1791 −6.20 −6.11 −2.54 −2.45 0.33 0.29 0.68 0.26 112.43(0.168) 123.93(0.046) 85.62(0.114) 86.83(0.098)
Notes: 1. In all models a full set of time dummies is included as covariates. m1 is the Arellano– Bond test for AR(1) in first differences and m2 is the Arellano–Bond test for AR(2) in first differences. Robust standard errors are in parentheses. *, ** and *** imply that coefficients are statistically significant at 10 per cent, 5 per cent and 1 per cent level, respectively. 2. The explanatory variables are FDI – a dummy variable taking the value 1 if there is at least one multinational enterprise in an industry, FS-E – the share of employment in multinational enterprise to total employment at industry level, lnSIZE – industry size measured by the logarithm of the total number of employees, MES – a measure of minimum efficient scale at industry level, HHI – the Herfindahl index of industry concentration, IGROW – industry annual growth rate, AGE – average age of enterprise in an industry.
but statistically insignificant and the second term negative and statistically significant. These results cast doubts on the u-shaped relationship between the net local entry rate and MNEs’ presence posited by theoretical models. In fact, they confirm our previous findings as the inverted u-shaped relationship indicates that initially MNEs promote entrepreneurship in the
Entrepreneurship and inward FDI in Portugal
173
Portuguese economy (assessed by a positive rate of local entry) but, as the MNEs’ weight on the industry structure increases, its positive effect vanishes, leading to the exit of a number of local enterprises that are not entirely replaced by start-up enterprises. These results, however, must be interpreted with caution, for it does not necessarily mean that FDI did not modify the type and characteristics of enterprises that comprise an industry. As FDI may affect local entrepreneurship in complex ways, the insignificant effect of the net entry rate might hide industrial restructuring through the replacement of inefficient enterprises by more efficient ones. Looking at the services industry and restricting FS_E to a linear impact, we find it having a negative and statistically significant effect on the net local entry rate. The inclusion of a quadratic term of FS_E renders both coefficients associated to MNEs’ presence negative and statistically insignificant, corroborating the finding that a u-shaped relationship between MNEs and local entrepreneurship seems to be inappropriate in terms of describing the Portuguese experience. With respect to the traditional determinants of entry, we find that industry growth and industry concentration are the main factors shaping entry behaviour, despite their opposite influences on the local net entry rate. In the manufacturing case, we find also a size of industry effect, indicating that local enterprises tend to enter industries populated by a medium or large number of enterprises. This suggests that local entrepreneurs show a weak capacity to promote changes in the industry structure and to be innovative in terms of the choice of industries to enter. To test the robustness of our results, we re-estimate the entry model by applying the semi-parametric Kernel estimator to the entire sample and to subsamples of our dataset. Figure 8.1 displays the estimated slope and Portuguese services industry
Net entry rate
Net entry rate
Portuguese manufacturing industry
0
0.2 0.4 0.6 0.8 1 MNEs’ presence at industry level (FS_E)
Figure 8.1
0
0.2 0.4 0.6 0.8 1 MNEs’ presence at industry level (FS_E)
Relationship between net local entry rate and MNEs’ presence for the Portuguese manufacturing and services industry
174
SMEs in a globalised world
shaping of the local entrepreneurship–FDI relationship using the predicted values of local net entry rate (purged from other covariates effects) for the manufacturing and services industries and the intensity of MNEs presence at industry level (FS_E). As can be seen, we fail again to find a u-shaped relationship between MNEs’ presence at industry level and local entrepreneurship for both manufacturing and services. More interestingly, the semi-parametric estimates suggest that the shape of the relationship under investigation is better described by a semi-inverted u-shaped relationship indicating that initially MNEs may stimulate would be-entrepreneurs but this positive effect (if any) is quickly vanished as the number and scale of MNEs increase at the industry level. The lack of evidence confirming a u-shaped relationship in the Portuguese context persuades us to look at relevant subsamples of the dataset. For that, we define different groups of industries, which are relatively homogeneous in terms of market structures, the formation of technological skills, and the sources and directions on technical change. The impact of MNEs on local enterprises’ innovative capacity is unlikely to be uniformly distributed among industries. On the contrary, technological knowledge (in general, and that brought by MNEs to the host economy, in particular) tend to be generally applicable and easily reproducible but specific to enterprises, cumulative in development and varied amongst industries (Pavitt, 1984). Therefore, differences in the intensity of MNEs’ presence among industries may have the same effect on entry if industries are dissimilar with respect to sources and direction of innovation and entrepreneurship. According to Pavitt (1984), we divide manufacturing industries into economies of scale, traditional, high-tech and specialized industries. Figure 8.2 shows the semi-parametric estimates of the local entrepreneurship–FDI relationship for the defined groups of industries. Comparing results for the four groups of industries, we find very dissimilar shapes for the local entrepreneurship–FDI relationship. We should, however, bear in mind that for higher values of MNEs presence at industry level, the estimates were obtained with less confidence than for other values as the distribution of observations for higher values of FS_E is relatively small. The specialized industries yield the less clear-cut estimates, making it very difficult to identify the slope of the relationship under scrutiny. In turn, economies of scale industries depict a slightly u-shaped relationship for small and medium values of FS_E, while traditional industries appear to not benefit from the presence and scale of MNEs at industry level. In these industries and for a significant range of values for FS_E, the competition effect of MNEs seems to systematically outweigh any positive spillover effect on local entrepreneurship.
Entrepreneurship and inward FDI in Portugal
175
Specialized industries
Net entry rate
Net entry rate
Traditional industries
0
0
0.2 0.4 0.6 0.8 1 MNEs’ presence at industry level (FS_E)
0.2 0.4 0.6 0.8 1 MNEs’ presence at industry level (FS_E)
High-tech industries
Net entry rate
Net entry rate
Economies of scale industries
0
0.2 0.4 0.6 0.8 1 MNEs’ presence at industry level (FS_E)
Figure 8.2
0
0.2 0.4 0.6 0.8 1 MNEs’ presence at industry level (FS_E)
Relationship between net local entry rate and MNEs’ presence for four groups of manufacturing industries
On the other hand, the group of high-tech industries seems to provide us with an interesting case. Excluding very small or large values of FS_E, the estimates of the local entrepreneurship–FDI relationship in those industries appear to provide empirical evidence supporting a u-shaped relationship, suggesting that the competition effect of MNEs outweighs any positive linkage effect on local entrepreneurship up until a certain threshold level. Once this threshold level is attained, one can expect a positive entry incentive in high-tech industries for local would be-entrepreneurs. The impact of foreign presence on local entrepreneurial activity may also vary according to size of enterprise. One important stylized fact about entry is that entrants are of smaller size than incumbent enterprises (Geroski, 1995), contributing to the pattern of right-skewness of the distribution of enterprise size observed in almost all industries. To assess whether there is a significant difference in the impact of MNE intensity on local net entry rates by entry sizes, we re-estimate Equation 8.1 computing the dependent variable for three distinct groups of enterprises (small, medium and large enterprises). Size of enterprise was measured
176
SMEs in a globalised world Medium manufacturing firms
Net entry rate
Net entry rate
Small manufacturing firms
0
0.2 0.4 0.6 0.8 1 MNEs’ presence at industry level (FS_E)
0
0.2 0.4 0.6 0.8 1 MNEs’ presence at industry level (FS_E)
Net entry rate
Large manufacturing firms
0
Figure 8.3
0.2 0.4 0.6 0.8 1 MNEs’ presence at industry level (FS_E)
Relationship between net local entry rate and MNEs’ presence for different sizes of enterprises
by the total number of employees. In particular, we define a small enterprise (entrant or exiter) if the total number of employees is less than 50, a medium enterprise is observed when the total number of employees lies between 50 and 249, and enterprises with 250 or more employees are classified as large enterprises. The semi-parametric estimation results on local entrepreneurial activity by type of enterprise are depicted in Figure 8.3. Again, we find very dissimilar shapes and slopes for the local entrepreneurship–FDI relationship for different types of enterprise. For small enterprises, the semi-parametric estimates suggest that the cumulative presence of MNEs do not endorse the development of local entrepreneurship. In the case of large enterprises, the upward slope of the estimated local entrepreneurship–FDI relationship suggests that the linkage effect prevails over any potential competitive effect. In any case, there is no evidence supporting a u-shaped relationship. Linking these findings with the distribution of size of enterprise in the Portuguese manufacturing industry (which display a predominance of SMEs) we might reinforce our previous conclusion that the role of MNEs in assisting Portuguese entrepreneurial
Entrepreneurship and inward FDI in Portugal
177
activity has been somewhat weak. The Portuguese SME sector appears not to take advantage from the competition challenge brought by MNEs.
CONCLUDING REMARKS MNEs are increasingly thought to be important for assisting and promoting the development of host economy enterprises. Theoretical models show that the entry and exit of local enterprises is governed by a competition and a spillover effect associated to the number and scale of MNEs. The competition effect first dominates but is gradually compensated by positive spillover effects. These two main economic countervailing forces at work shape the relationship between MNEs and local entrepreneurial activity and, therefore, show how local entrepreneurs respond to the competitive challenge and the opportunities created by MNEs in order to contribute for long-term economic revitalization of the local economy. Using enterprise level panel data for the manufacturing and services industries in Portugal and applying parametric and semi-parametric econometric techniques, our analysis does not confirm a positive relationship between MNEs and local firms (either manufacturing or services), as posited by theoretical models. We only found support for a u-shaped relationship in the case of high-tech industries, which seems to lend support to policies that promote the entry of MNEs in these industries. For some cases, the estimates suggest a linear relationship but, more importantly, a negative impact of FDI on local entrepreneurial activity. This is the case for the relationship between MNEs and the local entry rate on the SME sector. The Portuguese SME sector appears to respond negatively to the competitive challenge brought by MNEs. The number of entrepreneurs that recognize their inability to benefit from the challenges of globalization outweigh those that see the presence of MNEs as a potential channel to access new markets and growth. This suggests that the presence of MNEs constrains significantly the evolution of the Portuguese SME sector by creating an unfavourable competitive environment, leading some small and medium enterprises to choose exiting as the best option. Important policy implications emerge from these results. First, policies intended to attract MNEs should account for local industry-specificities and for the positive changes that can be induced in the industry structure in order to maximize welfare gains from FDI in general, and to assist local entrepreneurial activity, in particular. Second, the Portuguese experience suggests that FDI promotion policies per se could be an ineffective policy initiative. These policies should be complemented with other policies that: (a) focus on improving local enterprise characteristics and do not give
178
SMEs in a globalised world
preferential treatment to MNEs; and (b) enhance the interaction between local enterprises and MNEs, creating cooperative linkages among them. In motivating local enterprises to set up long-term cooperation with MNEs, policymakers should, nonetheless, pay attention to the greater competition MNEs may induce in both product and factor markets. Strong MNEs’ embeddedness may provoke the exit of local enterprises operating in other sectors due to shortage in the market factor, leading to distortions in the industrial structure and more dependency from MNEs. Therefore, policies should assist local entrepreneurs by helping them to gain the appropriate knowledge in order to benefit from the opportunities created by MNEs and, hence, to better respond to the challenges of globalization and industrial restructuring across countries. Noticeably, enterprises are not all equal in terms of their resources and competencies and, consequently, the desirable effects of a public policy may arise differently among local enterprises. In closing, it is worth noting that the weak evidence on the positive effects of MNEs on local entrepreneurial activity, measured by net entry rates, found in Portugal might conceal relevant impacts in terms of industrial restructuring, namely through the replacement of inefficient enterprises by more efficient ones. To disclose the nature and importance of this restructuring process induced by MNEs we have to analyse separately the intensity of entry and exit and the entrants and exiters’ characteristics. This point is beyond the scope of this chapter but it clearly deserves further attention in future research.
NOTES 1. The authors wish to thank the Portuguese Foundation for Science and Technology (FCT) for funding the project ‘Firm’s heterogeneous capabilities and industrial dynamics’ under research grant POCI/EGE/56937/2004 (partially funded by FEDER). 2. The Herfindahl index is a measure of market concentration that varies between zero (minimum concentration, maximum competition) and one (maximum concentration, minimum competition). 3. Unfortunately, information on the market share of all enterprises in an industry is not available. Nonetheless, the employment share of enterprises should serve as a good proxy for industry concentration given the well-documented correlation among the measures of enterprise size.
REFERENCES Acs, Z.J. and D.B. Audretsch (1989), ‘Small-firm entry in US manufacturing’, Economica, 56 (222), 255–65.
Entrepreneurship and inward FDI in Portugal
179
Adler, N. and N. Hashai (2007), ‘Knowledge flows and the modeling of the multinational enterprise’, Journal of International Business Studies, 38 (4), 639–57. Aitken, B. and A. Harrison (1999), ‘Do domestic firms benefit from direct foreign investment? Evidence from Venezuela’, American Economic Review, 89 (3), 605–18. Almeida, P. and A. Phene (2004), ‘Subsidiaries and knowledge creation: the influence of the MNC and host country on innovation’, Strategic Management Journal, 25 (89), 847–64. Arellano, M. and S. Bond (1991), ‘Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations’, The Review of Economic Studies, 58, 277–97. Barbosa, N. and H. Louri (2002), ‘On the determinants of multinationals’ ownership preferences: evidence from Greece and Portugal’, International Journal of Industrial Organization, 20 (4), 493–515. Barbosa, N. and H. Louri (2005), ‘Corporate performance: does ownership matter? A Comparison of foreign – and domestic – owned firms in Greece and Portugal’, Review of Industrial Organization, 27 (1), 73–102. Barbosa, N., P. Guimarães and D. Woodward (2004), ‘Foreign firms entry in an open economy: the case of Portugal’, Applied Economics, 36 (5), 465–472. Barrios, S. and E. Strobl (2002), ‘Foreign direct investment and productivity spillovers: evidence from the Spanish experience’, Weltwirtchaftliches Archiv, 138, 459–81. Barrios, S., H. Görg and E. Strobl (2005), ‘Foreign direct investment, competition and industrial development in the host country’, European Economic Review, 49 (7), 1761–84. Barry, F. and J. Bradley (1997), ‘FDI and trade: the irish host-country experience’, The Economic Journal, 107 (445), 1798–811. Cabral, L. (2007), ‘Small firms in Portugal: a selective survey of stylized facts, economic analysis, and policy implications’, Portuguese Economic Journal, 6 (1), 65–88. Cantwell, J. and L. Piscitello (2002), ‘The location of technological activities of MNCs in European regions: the role of spillovers and local competencies’, Journal of International Management, 8 (1), 69–96. De Backer, K. and L. Sleuwaegen (2003), ‘Does foreign direct investments crowd out domestic entrepreneurship?’, Review of Industrial Organization, 22 (1), 67–84. Driffield, N. and J.H. Love (2006), ‘Does the motivation for foreign direct investment affect productivity spillovers to the local sector?’, Applied Economics Quarterly, 52 (1), 3–28. Flores, R., M. Fontoura and R. Santos (2007), ‘Foreign direct investment and spillovers: additional lessons from a country study’, The European Journal of Development Research, 19 (3), 372–90. Fotopoulos, G. and N. Spence (1999), ‘Net entry behaviour in Greek manufacturing: consumer, intermediate and capital goods industries’, International Journal of Industrial Organization, 17 (8), 1219–30. Geroski, P. (1995), ‘What do we know about entry?’, International Journal of Industrial Organization, 13 (4), 421–40. Geroski, Paul (1991), Market Dynamics and Entry, Oxford: Blackwell. Girma, S., D. Greenaway and K. Wakelin (2001), ‘Who benefits from foreign direct investment in the UK?’, Scottish Journal of Political Economy, 48 (2), 119–33.
180
SMEs in a globalised world
Görg, H. and D. Greenaway (2004), ‘Much ado about nothing? Do domestic firms really benefit from foreign investment?’, World Bank Research Observer, 19 (2), 171–97. Görg, H. and E. Strobl (2005), ‘Foreign direct investment and local economic development: beyond productivity spillovers’, in T.H. Moran, E.M. Graham and M. Blomström (eds), Does Foreign Direct Investment Promote Development?, Washington, DC: Institute for International Economics, pp. 137–57. Härdle, W. (1991), Smoothing Techniques, With Implementation in S, New York: Springer-Verlag. Härdle, W., M. Müller, S. Sperlich and A. Werwatz (2004), Nonparametric and Semiparametric Models, New York: Springer-Verlag. Kathuria, V. (2000), ‘Productivity spillovers from technology transfer to Indian manufacturing firms’, Journal of International Development, 12 (3), 343–69. Lin, P. and K. Saggi (2005), ‘Multinational firms and backward linkages: a critical survey and a simple model’, in T.H. Moran, E.M. Graham and M. Blomström (eds), Does Foreign Direct Investment Promote Development?, Washington, DC: Institute for International Economics, pp. 159–74. Markusen, J. and A. Venables (1999), ‘Foreign direct investment as a catalyst for industrial development’, European Economic Review, 43 (2), 335–56. Mata, J. (1993), ‘Entry and type of entrant: evidence from Portugal’, International Journal of Industrial Organization, 11 (1), 101–22. Mata, J. and P. Portugal (2004), ‘Patterns of entry, post-entry growth and survival’, Small Business Economics, 22 (3–4), 283–98. McKeon, H., K. Johnston and C. Henry (2004), ‘Multinational companies as a source of entrepreneurial learning: examples from the IT sector in Ireland’, Education Training, 46 (8/9), 433–43. Pavitt, K. (1984), ‘Sectoral patterns of technical change: towards a taxonomy and a theory’, Research Policy, 13 (6), 343–73. Rivera-Batiz, F.L. and L.A. Rivera-Batiz (1990), ‘The effects of direct foreign investment in the presence of increasing returns due to specialization’, Journal of Development Economics, 34 (1–2), 287–307. Robinson, P. (1988), ‘Root-n-consistent semiparametric regression’, Econometrica, 56 (4), 931–54. Rodríguez-Clare, A. (1996), ‘Multinationals, linkages, and economic development’, American Economic Review, 86 (4), 852–73. Salgado-Ugarte, I.H., M. Shimizu and T. Taniuchi (1994) ‘snp6: exploring the shape of univariate data using kernel density estimators’, Stata Technical Bulletin, 3 (16), 8–19. Scott, D.W. (1992), Multivariate Density Estimation: Theory, Practice, and Visualization, New York: John Wiley and Sons.
9.
The large leader firm: good or bad? A case study of a leader firm–supplier relationship Helen McGrath and David Jacobson
INTRODUCTION This chapter begins with a review of the literatures on industrial districts and leader firms. Over the period since the 1980s many ‘traditional’ industrial districts of large numbers of small firms in one location, specializing in the production of different parts of the same product, have undergone significant structural change. In some cases they have simply gone into decline as a result of loss of competitive advantage to cheaper locations of production in the Far East (for example, the textile industrial district of Como). In other cases, as a result of market expansion and integration and the forces of globalization in general, they have been displaced, the small firms going out of existence, being replaced by a small number of large firms (for example, the eyewear districts of Belluno, in the Veneto region). This raises the question of whether, in the context of such forces of globalization (lower costs of production, market expansion and integration) the possibility exists of positive associations between small firms from within small firm networks, and large, leader firms. Given that, as will be shown, there is indeed evidence of such positive associations, we focus on a specific example and question the extent to which it corroborates the international evidence. The particular case examined in this chapter is that of the fish processing industry in North Dublin, Ireland. There is evidence of industrial districtlike structure and behaviour in this sector. One of the firms in this district formed a close, trust-based relationship with a large supermarket chain, the former pinning its strategy to the latter as a leader firm. The result has been significant growth of the fish processing firm which has emerged as a leading firm in the industry nationwide. The relationship between a small, start-up fish processing firm and a large supermarket company in this case proved to be mutually beneficial. The chapter provides suggestions as to 181
182
SMEs in a globalised world
what factors might have contributed to success in this case, and might have been absent from cases in which the forces of globalization – leading to an association between a large, leading firm and SMEs – had more negative consequences for SMEs.
THE EVOLVING INDUSTRIAL DISTRICT Industrial districts, as described in a now vast literature, consist of a large number of spatially concentrated small firms all involved in the production of the same or closely related products. Among the main characteristics of industrial districts is that these small firms both cooperate and compete with one another, the competition tempered by close social relationships among the owners of the firms (Brusco, 1982; Piore and Sabel, 1984; Becattini, 1991; Dei Ottati, 1991). Even among situations remaining consistent with this definition, a wide variety of types of industrial district existed, of different sizes, different degrees of industrial and commercial autonomy, different levels of technology and in more and less developed economies (Paniccia, 1998). More radically, in recent years some industrial districts have more or less ceased to exist, while others have changed to such an extent that, even though they may maintain some of the key elements, they are no longer consistent with the canonical definition.1 Recent developments have included the decline of the clothing and textile district in Como, employment falling from over 36 000 at the beginning of the 1990s to less than 26 000 in 2003 and output from a high of over €2.9 billion in the mid-1990s to less than €1.9 billion in 20032 (Alberti, 2006). Alberti (2006) identifies intensified foreign competition in the context of increased globalization as a main cause of the decline. Among the local responses have been switching linkages from local to international value chains (relocation) and changing product focus to high-quality niches (reconversion). He concludes that this relocation and reconversion may enable a reduced cluster of firms to survive but that the historical industrial district of Como is on a downward trend. Another Italian industrial district with a focus on clothing and textiles that now is in decline is the Vibrata–Tordini–Vomano (VTV) district (Sammarra and Belussi, 2006, p. 559). Again foreign, and particularly Asian competition, led to structural decline in the 1990s. In VTV the prospects appear to be even worse than in Como. Sammara and Belussi (2006) identify the VTV strategy as one of ‘replicative relocation’. By this they mean the types of activities still remaining within the district are not particularly different from those relocated abroad. They proceed to argue that although allowing some of the larger firms to maintain their
A case study of a leader firm–supplier relationship
183
competitiveness in the medium term, the general survival of the overall district is put into question. This may enable some of the larger local firms ‘to preserve their competitiveness in the medium term, [but] the survival of the overall district is put into question’. Specifically referring to the ‘challenge from China’, Russo (2004) analyses the contradictions in the recent dynamic in the ceramic tile district of Sassuolo. The main element of the challenge is the ability of the Chinese capital goods industry to produce the tile manufacturing equipment. This had been a key competitive advantage, along with the tile machine firms’ close relationship with the tile makers, of the Sassuolo district. In his description of Russo’s work, Sabel (2004) suggests that the recent developments may undermine both the capital goods sector at the heart of the district and the tile makers themselves. The transformation of industrial districts is not confined to those in Italy. An Irish industrial district – that of furniture in County Monaghan (Jacobson and Mottiar, 1999) – also appears to be undergoing substantial change. This was the first, and so far only, Irish industrial district to be identified in the academic literature. It was shown to have most of the characteristics of Emilian districts. More recently some of the leading firms in the district have been shown to have changed their production systems in such a way as to significantly reduce their local industrial embeddedness. First, the largest firm in the district, Coyles, supported by the Irish development agency Enterprise Ireland, invested some €4.5 million at the end of the 1990s in re-engineering its manufacturing facility towards modular furniture production. This inevitably involved a reduction in local linkages, both upstream and downstream. Since then, similar changes, with similar consequences, have taken place in other leading firms in the district (Heanue and Jacobson, 2008). Going beyond case studies, Heanue (2007) statistically analyses the Irish furniture industry, showing that the attractiveness of County Monaghan as a location for furniture firms has declined over the decade to 2006. Whether the absence of industrial district-type embeddedness of firms in this location is a cause or effect of the reduced attractiveness as a location for furniture firms, it is clear that it is has less and less of the characteristics of the traditional industrial district. Industrial districts have evolved in different ways; the above examples are mostly negative but many others are positive. The sofa district of Matera in the Basilicata region of Italy has come to specialize in leather furniture. Beginning as a typical district based on small firms, primarily artisan-owned, using tacit knowledge and skills, it has evolved, under the leadership of Natuzzi, into a world-leading producer of leather sofas (Albino et al., 1999). Among Natuzzi’s innovations was the introduction
184
SMEs in a globalised world
of computer aided design (CAD) and modern quality control systems. Through its links to the export market, Natuzzi was able to provide information about international market demand. This also gave it the leverage to direct the introduction of its innovations among its local suppliers. Here there is clear evidence of a large, leader firm’s relationships with small, spatially proximate firms advantaging both. A similar increase in competitiveness led by leader firms in an Italian region is reported by Camuffo (2003). The small-firm industrial districts manufacturing eyewear (spectacle frames) in the Veneto region went into crisis in the second half of the 1990s. The competitive advantage was in flexible production organization, where cost is most important, rather than marketing, distribution, product innovation, design and fashion leadership, where cost is less important. As a result low-cost production in China proved to be more competitive. Globalization resulted in the shift of production, at least of the low-cost parts of the production process, to China and other low-cost countries. Small-firm industrial districts that focused on production to the exclusion of marketing and product innovation were unable to compete. This might have been the end of the eyewear industry in Veneto, but a small number of larger firms were able to grow and compete in world markets. From the 1980s they began to take control of some of the small, local suppliers; one acquired a wholesale distributor, another a number of retail chains. Moving closer to their final customers provided information about changes in demand for their products. With their increasing vertical integration, these companies gained from economies of scale and were able to introduce new production technology. By the early 2000s four of the top ten firms in the world in this industry had their home bases in Veneto, including Luxottica, the world number one. Some of their suppliers have also grown, under the leadership of firms like Luxottica (Camuffo, 2003). In these latter examples, the emergence of large, leader firms, though altering the dynamic of some of the traditional small firm industrial networks, also maintains or increases the competitiveness of their industries in their respective regions. It has been suggested that such large firms can ‘organize production among groups of smaller firms, introduce technological innovation and expand existing markets’ (Lazerson and Lorenzoni, 1999, p. 235). The role played by leader firms in orchestrating network relations can, however, have both positive and negative ramifications for the firms operating in industrial districts. On the one hand, leader firms provide ‘the preconditions for industrial districts through the transfer of technology and skills to an area which ultimately provide the foundation for the creation of small firms’ (Lazerson and Lorenzoni, 1999, p. 240). Their exploration of commercial avenues and investment in R&D link the
A case study of a leader firm–supplier relationship
185
leader firm both to distant and to local actors, which places them in a strategic position in order to be able to respond promptly to external market demand, whilst at the same time realigning the productive resource of the less sensitively located actors. On the other hand the leader firm can introduce oligopoly (or oligopsony) power that may threaten the typical model of cooperative competition present in the Italian industrial districts (Harrison, 1994a; 1994b). Through their domination and standardisation of supplier linkages, power structures are such that non-reciprocal arrangements replace what were once symbiotic exchanges (Jacobson and O’Sullivan, 1994; Boschma and Lambooy, 2002). We turn now to some consideration of the different circumstances resulting in positive and negative impacts of leader firms.
LEADER FIRMS: THE MECHANISMS The preferential relations that a leader firm develops with its supplier or group of suppliers can be influenced by whether a low or high opportunistic rent is associated with the transaction (Dei Ottati, 1991). Where the threat of (financial return from) opportunistic behaviour is low, local customs, social trust or norms of economic behaviour may provide sufficient means of coordination. Using Grandori’s (1997) terminology this type of organic interfirm connection may be referred to as a ‘social’ network (McGrath, 2006). Repeated cooperative interaction (that leads to the establishment of reputation and trust) can also safeguard against opportunistic behaviour. In cases where the threat of such behaviour is high, Williamson (1985) for example, would argue that there is a need for vertical integration so as to hedge against such opportunism (reduce transaction costs). Vertical integration, however, carries its own set of difficulties. In relation to management systems, Rugman and D’Cruz (2000) attribute failure of the vertically integrated firm to the multidivisional or M-form arrangement used in management systems. They suggest that it lends itself to internal rivalry among general managers and to intraorganizational rivalry in general. As an alternative form of coordination they suggest a process of disintegration into a partner-type network. This, they term the ‘flagship model’. At the centre of the leader firm concept is an organizational structure that opts for coordination integration3 rather than ownership integration. Essentially, it involves a process of disintegration rather than corporate vertical integration (McGrath, 2006). How interfirm relationships are constructed within this process of coordination integration enables us to determine whether leader firms are ‘good’ or ‘bad’ (that is, have positive or
186
SMEs in a globalised world
negative long-term impacts on the firms they lead, and in particular on the independence of those firms). Within the flagship model, suppliers develop collaborative relationships with the flagship firm that are based neither on arm’s-length market transactions nor on a short-term orientation with a single transaction focus. Instead, there is a collective long-term orientation that facilitates the development of long-term competitiveness. Through ‘strategic asymmetry’ the flagship firm exercises non-reciprocal control over network partners’ products, markets and capital investments as well as courses of action in the development of competencies and capabilities. In return, the flagship network member expects increased sales, access to technological advances and the benefits of the flagship’s brand image (Rugman and D’Cruz, 2000). The aspects of this model that may deem it a ‘bad’ leader firm arrangement are unilateral rather than bilateral controls and the degree to which suppliers may over-rely on the flagship/leader firm for custom (particularly if exclusivity is a feature, that is where the leader firm is the sole customer of the suppliers). To categorize such leader firm activity we can again borrow from Grandori (1997) and call such leader firm–supplier connectivity a ‘bureaucratic’ network. Bureaucratic networks, although based on formal and explicit contracts can, in some cases, involve relational contracting that combines contract with unwritten rules and safeguards (Grandori, 1997). How different coordination mechanisms emerge or evolve is dependent on the type of interdependence between firms. Applying this rationale to a leader firm–supplier relationship, an assessment must be made as to whether the interdependence is intensive, sequential and/or reciprocal (Grandori, 1997). The type of interdependence and the network type (social or bureaucratic) dictate the nature of the leader firm–supplier relationship, in other words whether it is ‘good’ or ‘bad’ (McGrath, 2006, Chapter 7). The different types of interdependence to which we have referred deserve some further explanation.4 Where an intensive (innovative) interdependence exists, differentiated professional know-how is jointly applied to a common problem. In such cases trust-based relationships are nurtured facilitating ‘learning by interacting’ (Carbonara, 2004) and the diffusion of knowledge (Albino et al., 1999). It follows that the leader firm is accorded a central role in the innovation processes. Acting as a bridge between internal and external knowledge the leader firm channels new knowledge from its regional, national and often global links back to district-based relatively small suppliers (Boschma and Lambooy, 2002). Where the interdependence is sequential (a link between two activities for which the output of an activity A is the input of activity B) the benefits of the relationship to the supplier are limited, particularly if ‘bureaucratic’ means of
A case study of a leader firm–supplier relationship
187
coordination (one way hierarchical subcontracting) are employed. In such cases the lack of mutual adjustment means that the benefits to suppliers remain purely transactional. Reciprocal interdependence (like sequential, except it is not possible for the output of A to be produced without certain inputs from B such as elements of know-how or other resources) whether present in bureaucratic or social networks is characterized by flexibility. In such instances the potential for relational contracting, mutual monitoring and liaison activities is great. It is not possible to divorce the leader firm from its particular context, in particular the relationships it has with its suppliers, its development strategies and so on. Therefore the merits or demerits of leader firms in local economies are a matter for empirical exploration.
CASE STUDY: THE EVOLUTION OF A LEADER FIRM–SUPPLIER RELATIONSHIP Background The leader firm in question is Superquinn (retail buyer) and one of its suppliers, Oceanpath (a small5 fish processor). The headquarters of the former firm are located in Sutton (north County Dublin) and the processing plant of the latter firm is located in Howth (north County Dublin). The history of interaction between the founders of Oceanpath and individuals in Superquinn offers a great deal of insight into how this leader firm–supplier relationship has been constructed and how it operates on a day-to-day basis. Furthermore, elements of Oceanpath’s contextual environment6 have reinforced the leader firm–supplier collaborative processes due to the general climate of cooperative competition that prevails among fish processors in Howth and the ‘collective action’ (Becattini, 1991; Sengenberger and Pyke, 1992; Schmitz, 2000) approaches to problem solving (for example in relation to sourcing of inputs) to which Oceanpath has become accustomed. Oceanpath has been Superquinn’s sole supplier of fresh fish since 1995 (when Oceanpath was founded). However, there is a history of recurrent contracting and interpersonal links that predates the establishment of Oceanpath. Prior to 1995 Superquinn sourced fresh fish from a number of different processors but primarily from another Howth-based processor. From the late 1960s to the end of 1994, Superquinn bought fish from this processor. For almost 30 years a manager of this processor, who was to become one of Oceanpath’s directors, handled the Superquinn account. The frequent, often face-to-face encounters and repeated exchanges
188
SMEs in a globalised world
between himself and the Superquinn buyers facilitated the development of process-based trust, that is trust developed over a period of time (Lorenzen, 2002) and through ‘imputation from outcomes of prior exchanges [which] provides data for current or future exchanges’ (Dwivedi et al., 2003, p. 96). It was in 1971 during a fish blockade that commitment on the part of the original fish supplier and in particular by one of its shareholders/ managers (co-founder of Oceanpath) was first demonstrated. At the time, supermarkets in Ireland were unable to source fish. However, through this man’s contacts an arrangement was made for Superquinn that enabled it to be the only supermarket in the country with a supply of fresh fish. The director’s demonstration of commitment, reliability and probity led to Superquinn’s trust in him (McGrath, 2006). Throughout the almost 30 year supplier–buyer relationship, exchanges were market-based (that is on a transaction by transaction basis) and Superquinn practised multiple sourcing so as to hedge against inaccurate pricing or opportunistic behaviour. A turning point in Superquinn’s approach to suppliers can be traced to its involvement in the European Coca Cola Research Group from 1987 to 1992 (Kelly, 2002). Through interaction with European counterparts Superquinn learned that supplier– retailer collaboration (SRC), particularly with regard to fresh foods where the retailer is the brand (as opposed to manufacturer branded goods), could improve performance significantly. Superquinn’s response to this information was to initiate a supplier partnership with a number of fresh food suppliers, that is, suppliers of beef, chicken, eggs, fish, cheese and vegetables. Superquinn committed itself to single sourcing – an expression of trust in customer–supplier relationships (Harrison, 1994a) – provided the supplier demonstrated commitment, openness and trust (COT). Commitment meant that both sides would fully commit all resources to an agreed change programme within a given time frame. Openness on the supplier side meant Superquinn offering open book costings and transparency in all its dealings with suppliers, and the equivalent on the other side. This facilitated the evolution of dyadic trust. Superquinn’s rationale behind the development of collaborative or cooperative relationships with suppliers mimics the strategies of leader firms in the study of firms and networks in the furniture industrial districts of Bari-Matera, Forli, Pistoia and Udien where leader firms established collaborative relationships only with suppliers or subcontractors that ‘play a strategic role in the value chain and/or have a strategic complementary competence’ (Carbonara, 2002, pp. 242–3). In this sense Superquinn – as the leader – is determining the strategy. It is initiating the changes and determining the rules of a new game.
A case study of a leader firm–supplier relationship
189
Due to a number of factors, such as unwillingness to commit to change, reluctance to operate on the basis of open book and in particular its slow rate of development and use of traditional production techniques, a partnership with the original fish supplier was not entered into. Entrepreneurial vision on the part of three of its shareholders spurred them to sell their shares in the company and form a rival firm – Oceanpath – with the express purpose of entering into partnership with Superquinn. Like the leader firms of industrial districts, Superquinn’s supplier selection process was not purely based on price but also on reliability, flexibility and innovation capability (Carbonara, 2002). Oceanpath’s youthfulness was perceived by Superquinn to be a positive attribute in that the firm could be moulded to suit Superquinn’s requirements in terms of supply and quality (McGrath, 2006). Partnership The partnership is an informal one in that no written contract exists to bind the firms as partners. Although the relationship is coordinated on the basis of structured or conscious mechanisms, for example, mutual monitoring and liaison activities, as described by Grandori (1997), it may be classed as a ‘social’ rather than ‘bureaucratic’ network. The interdependency between the two firms can be described as sequential and intensive. The relationship is, however, to some extent reciprocal. It bears many of the features of a symmetrical partnership, that is, for example, sharing of resources, joint decisionmaking and joint marketing. We have already indicated that successful leader firm–supplier relationships are those which are collaborative, built upon trust and are subject to some kind of bilateral controls (formal or informal). It follows that in order to assess whether the relationship between Superquinn and Oceanpath constitutes a ‘good’ leader firm–supplier arrangement consideration must be given to these elements. Our assessment is categorized under two headings: collaboration and dependency. To evaluate the degree of collaboration and dependency we have applied Carbonara’s (2002) parameters. For collaboration, she suggests examining the frequency of communication, resource sharing and information sharing about processes and products as well as codesign practices. Carbonara (2002) uses a binary system of evaluating these parameters (that is, high or low). Where all the indicators are given a high value then there is a high degree of collaboration. For dependency, the parameters are the degree of exclusivity of the relationship and the degree of dependence on technology and knowledge. Similarly, where there is a high value attaching to all then a high degree of dependency is implied. Table 9.1 illustrates the degree to which the relationship between
190
Table 9.1
SMEs in a globalised world
Collaboration between leader firm and supplier: Superquinn and Oceanpath
Parameters for degree of collaboration
Value
Between Superquinn and Oceanpath
Frequency of communication Resource sharing
High
●
High
● ● ● ●
Information sharing about processes and products
High
● ● ● ●
Codesign practices
High
● ● ● ●
Daily contact – telephone, email and face-to-face interaction Personnel Financial Buying power Office space and equipment Market information – via access to SAP ‘Bridge’ to information Quality assurance Traceability New product development In-store product presentation Joint marketing Launch of new quality standards
Oceanpath and Superquinn is collaborative and Table 9.2 indicates the degree of dependency. Elaboration on each of the points will shed further light on the nature of the relationship. Frequency of communication The personal and frequently face-to-face contact between Oceanpath’s business relations manager and the Superquinn buyer, in terms of knowledge transfer processes, can be described as a medium with a high degree of richness, that is, it facilitates equivocality reduction (Albino et al., 1999). The frequency of communication is facilitated by the fact that Oceanpath’s directors have swipe card access passes to Superquinn’s head offices in Sutton where they can meet with Superquinn personnel on both an informal and formal basis and access any documentation relating to the seafood category (as if they were a Superquinn employee). Resource sharing In terms of resource sharing, Superquinn has assigned personnel for comarketing and cogeneration of new production development ideas – meetings for which are held every two months. In 2002, Oceanpath built a new production facility at a cost of €3 million. This would not have been possible without Superquinn’s support. Oceanpath can avail of
A case study of a leader firm–supplier relationship
Table 9.2
191
Dependency between leader firm and supplier: Superquinn and Oceanpath
Parameters for degree of dependency
Value
Between Superquinn and Oceanpath
Degree of exclusivity of relationship
High/ Low
● ●
Degree of dependence on technology and knowledge
High
● ●
Degree of dependence for strategic development
High
●
Superquinn has no other supplier of fresh fish Oceanpath is free to supply to other retailers Superquinn is Oceanpath’s ‘bridge’ Superquinn needs Oceanpath’s expertise in seafood Superquinn’s strategic guidance led to Oceanpath’s status as top wet fish producers
Superquinn’s scale in terms of buying power which has enabled the firm to make cost savings on the purchase of non-input materials (for example, stationery and protective clothing) but also on input materials like sauce ingredients for value added products. Prior to the completion of the new production facility one of Oceanpath’s directors occupied an office in Superquinn’s nearby head offices in Sutton. For six months he acted as Superquinn’s buyer (while suitable candidates were being sought) without remuneration. At the same time, the Oceanpath manager was able to access Superquinn resources (as if he were an employee). The above provides evidence of reciprocal resource sharing. Information sharing about processes and products The leader firm can act as a valuable source of knowledge providing ‘information about the supplying markets, training on quality management, and production techniques, collaboration at the beginning of the order, visits to check the state of the order, etc.’ (Carbonara, 2002, p. 235). Where Superquinn has not had direct knowledge of a process or aspect of production it has acted as a gateway to that information for Oceanpath. Boschma and Lambooy (2002) class this type of leader firm as a ‘bridge’ leader firm. They suggest that bridge leader firms have external, sometimes global links. Followers benefit from these links by means of external knowledge that is transferred to them. Before Oceanpath installed a traceability system they examined the traceability system in place in the Carton Group’s chicken processing operations (based in County Cavan, Ireland) and adapted it
192
SMEs in a globalised world
for fish processing (Carton Group is another Superquinn supplier). Using Bender and Laestadius’s (2005, p. 8) terminology Oceanpath’s ‘innovation enabling capability’ lies in its ability to ‘transform available general knowledge and competence into plant, firm or task specific knowledge and competence’ (that is, its transformative capabilities). The way in which Oceanpath has been able to transform knowledge engendered in equipment and specific to another sector of industry (that is, chicken processing) and modify it for its own needs has enabled the creation of know-how and led to new skills and routines being acquired (McGrath, 2006). Superquinn, over a number of years and through a process of empirical learning (that is, learning by doing, learning by using and trial and error processes (Carbonara, 2004)) with its other perishable food partners, has developed its own in-house quality assurance/hygiene standard. This standard has resulted in the ‘codification’ of technical and organizational knowledge with regard to ‘quality’ procedures in production processes – the resulting ‘codebook’ is the Superquinn Hygiene Manual. Leader firms in Italian industrial districts, to control each aspect of production, have used this codification procedure. ‘By the codification process and the consequent higher knowledge transfer speed, the leader firm has been able to improve the knowledge communication and acquisition processes’ (Albino et al., 1999, p. 61). A more recent development has given Oceanpath access to market information through Superquinn’s SAP (business software) system. This is an example of both resource sharing and information sharing. As a valuable market tool, SAP can provide Oceanpath with data on real time product sales that can assist in determining what inputs to source, for example what type of fish to buy and which value added product lines to expand upon, reduce or withdraw. This information is also useful in any dealings Oceanpath has with other customers (retailers). Codesign practices Codesign practices can be interpreted as any joint approach to product design involving the two firms. Despite the fact that Superquinn is the brand on any value added meal produced for them by Oceanpath, Oceanpath has an input into its development and final presentation. Similarly, with regard to marketing of products, Oceanpath, along with the Superquinn marketing team, develop in-store marketing campaigns. There is therefore an intensive interdependency between the two firms, that is, ‘joint application of complementary resources to a common problem in an integrated way’ (Grandori, 1997, p. 902). The codevelopment of the IQS (Irish Quality Salmon) standard under the umbrella of QSP (BIM’s7 Quality Seafood Programme) reaffirms the claim that there is an intensive
A case study of a leader firm–supplier relationship
193
interdependency. Along with BIM, Superquinn and Oceanpath collaborated to launch a new standard for the packing and handling of farmed salmon which, as yet, has not been implemented by any other secondary processor or retailer. Degree of dependency Exclusivity is unilateral between Superquinn and Oceanpath. In other words, Oceanpath is Superquinn’s only supplier of seafood yet Oceanpath has other customers – among them, some of Superquinn’s competitors. Similarly, while Superquinn may obtain certain chicken products only from the Carton Group, Carton is free to sell to other retailers and, in fact, also have a strategic alliance with Musgraves.8 In Carbonara’s (2002) examination of eight leader firms the degree of exclusivity was high in all cases. However, where exclusivity evolves into monopsony – that is where the leader firm stipulates, or market conditions prescribe, that the supplier is to deal solely with them – dependence is not in the best interest of the supplier. Harrison (1994b) argues that the emergence of oligopoly power as exemplified in the case of leader firm, Benetton, its subcontractors and subsuppliers, is detrimental to cooperative competition in industrial districts. However, where exclusivity is as described in Superquinn and Oceanpath’s case, the leader firm encourages the supplier firm ‘to develop multiple relations with other final firms, rather than seeking to hold them hostage’ (Lazerson and Lorenzoni, 1999, p. 258). Through Oceanpath’s continuous relations with other customer firms (for example, SuperValu, EuroSpar and Dunnes Stores) it can avail of ‘many more points of contact and information than agents working within hierarchical organizations where information is often purposely limited’ (Lazerson and Lorenzoni, 1999, p. 258). In assessing the degree of dependence that the supplier has on the leader firm for technology and knowledge, Carbonara (2002) states that such technology and knowledge is owned by the leader firm. Superquinn has not directly provided technology but it has been the driving force behind Oceanpath’s use of new processes and automated production. With regard to Superquinn as a source of information, there is a high degree of dependence. This dependence is, to some extent, reciprocal. Superquinn needs Oceanpath’s knowledge of products and of the seafood sector as a whole to ensure that it is a ‘knowledgeable’ retailer (especially in light of the fact that its brand name is on Oceanpath-produced value added meals). For Oceanpath, knowledge in relation to consumer tastes/ demands, quality standards and organizational administration obtained from Superquinn, has enabled it to become one of the top wet fish processors in Ireland (McGrath, 2006). An additional proxy may therefore be proposed: dependence on leader firm for strategic development.
194
SMEs in a globalised world
In using Carbonara’s (2002) parameters to assess the degree of collaboration between Superquinn and Oceanpath we can conclude that the relationship is a collaborative one. The importance of collaboration or cooperation cannot be overstated. It is associated with knowledge spillovers, localized learning, improved innovative capacity and innovation processes that constitute systems of innovation (Maskell and Malmberg, 1996; Koschatsky, 1999; Capello, 2002; Garibaldo and Jacobson, 2005). Having established this, we can examine in more detail the changes that have taken place within Oceanpath since it has become aligned with Superquinn. Relationship Impact As outlined by Albino et al. (1999), the district leader often plays a relevant role regarding the introduction of a modification. They proceed to argue that ‘Production process automation, quality control, computer aided design are some of the activities that the leader firm tends to promote within its organization and to require from its suppliers’ (Albino et al., 1999, p. 58). In addition, leader firms can promote new product development and the introduction of organizational change for example in the form of planning processes and operation management. Changes in Oceanpath’s production process When Oceanpath commenced production in 1995 it did so using simple and largely manual techniques. The quality checks were of a very basic nature – purely sensory and without any probing. This has changed considerably since production first began. Now, a computer-based factory management system records the weight and temperature of all stock. Tracer tags are allocated to each box of fish. The tracer tag, if scanned, shows the temperature it has been stored at and where it has come from (that is, the name of the boat, cooperative, fish farm and so on). Although traceability has been a legal requirement since 1 January 2004, Oceanpath has had this system in place since 2002. Under the agreed change and development plan this was a Superquinn requirement. Once Oceanpath occupied its new facility in October 2002 its operation became much more automated. Prior to the move, salmon cutlets and skinless and boneless products for example, were produced manually. This process meant that filleters had to remove skin manually and pluck each bone by hand. At that time only 25 000 cutlets could be produced in a week. Since automation and the introduction of a skinning and boning machine, the number of cutlets produced in a week has increased to 120 000. The same numbers of filleters are employed now as when the
A case study of a leader firm–supplier relationship
195
process was manual, as they are required to prepare the fish for machine processing. A new packing machine has dramatically increased the firm’s output of pre-packed products. When packers had to cling-wrap products by hand approximately 1000 to 1200 packets were produced in a day. Since the introduction of the packing machine the same number of units can be produced in an hour. These improvements can be directly attributed to Oceanpath’s partnership with Superquinn. Superquinn required that output capacity be increased in order to cope with the demand of instore promotions. Rather than just stating its requirements, Superquinn lent its support to ensure the successful implementation of more automated processes by providing technical information via its other partners (for example its chicken partner, Carton Group). When questioned for this study on the standard and type of processing equipment that has been installed in Oceanpath, BIM representatives commented that it was unusual for a firm of Oceanpath’s size to have such high tech machinery. Related to changes in Oceanpath’s production process, is the introduction of quality standards. This has had implications not only for production but for internal organization also; employment of a quality assurance (QA) manager was a Superquinn stipulation. Accreditation to the British Retail Confederation (BRC) higher-level global food standard in July 2004 has superseded other quality assurance schemes such as Superquinn Hygiene that Oceanpath had previously operated under. Having a BRC certificate means that Oceanpath is approved to supply retailers like Sainsbury, Safeway and Tesco as well as Superquinn. As the only seafood processor in Ireland with this standard Oceanpath has acquired a distinct competitive advantage. Product changes In 1995, Oceanpath mainly produced white fish such as cod and haddock. In 1998 it introduced tuna and swordfish and more recently it has introduced strawberry grouper. Product changes have been influenced by changes in consumer taste, information on which is supplied by Superquinn market research. In addition to these changes, Oceanpath has developed a value added range requested by Superquinn. Through a process of trial and error, products are invented and then given a threemonth trial period on Superquinn shelves. Depending upon sales levels the product is continued or withdrawn. Organizational change Organizational change within Oceanpath can be attributed to its accreditation to quality assurance standards. The ‘codebooks’ in relation to
196
SMEs in a globalised world
quality practices specify certain action in relation to general management practices, personnel management, factory management, production management as well as transport and systems management. Benefits to Superquinn Like Oceanpath, Superquinn has benefited greatly from the relationship. It has wet fish counters in all of its 19 stores. As Superquinn has grown, so too has its ability to provide fresh fish. The category represents approximately 1.8 per cent to 2 per cent of Superquinn’s turnover. It therefore cannot justify a dedicated seafood buyer. However, without Oceanpath as a partner the seafood category alone would occupy 30 to 40 per cent of the Superquinn buyer’s time. As Oceanpath’s business relations manager does much of the work in relation to this category the Superquinn buyer is free to concentrate on his or her other categories (for example, bakery and fruit and vegetables). ‘It’s like having a buyer without having a buyer’ said the Superquinn buyer, referring to Oceanpath’s business relations manager (McGrath, 2006, p. 157). The partnership has enabled Superquinn to have control over standard and quality of production without having to vertically integrate. The Superquinn–Oceanpath partnership can be described as quasi-vertical integration (Langlois and Robertson, 1995). In the context of innovation, Superquinn has been able to add expertise and knowledge specific to the fish processing industry to its distinctive competencies in retailing and marketing. We can conclude that there is a high degree of coordination integration.
CONTEXTUAL ENVIRONMENT One of the main criticisms of the leader firm as a model of industrial development is that the exclusivity which is created between a number of firms in a group formation (that is, between the leader firm and its small network of suppliers) can cause the decline of an industrial district or sector due to the fact that knowledge is not widely diffused. Instead, knowledge diffusion occurs only between the leader firm and its core suppliers (Boschma and Lambooy, 2002; Coro and Grandinetti, 2001). The above is not the case between Oceanpath and other fish processing firms due to two factors: the set of prevailing conventions or ‘rules of the game’ by which firms and their agents conduct business; and the structure of the sector (Jacobson and McGrath, 2006; McGrath, 2006). The norms of economic behaviour are such that competitors share information (commercial) with one another in an informal manner; resources such as
A case study of a leader firm–supplier relationship
197
refrigeration space have been shared between firms in the past and inputs are sourced having regard to all (local) processors’ needs, demonstrating a collective approach to the sourcing of inputs. The structure of the sector is such that Oceanpath for example, in addition to its processing activities acts as wholesaler at the Dublin Fish Market. In doing so, it sells to another Howth (County Dublin) processor – a processor with whom they also compete against to supply to retail buyers.9 Due to the fact that Superquinn does not place any limitations on what companies Oceanpath sell to, and how they interact with either their competitors or customers, means that these conventions or ways of conducting business are not inhibited. The benefits of the leader firm–supplier relationship to Oceanpath’s customers are the improved quality and standard of product received. For other processors, there are some indirect benefits for example via informal interaction at the fish markets where knowledge is shared (McGrath, 2006). There appears to be nothing about the leader firm’s relationship with its supplier that undermines the industrial districtlike characteristics of fish processing in north Dublin. On the contrary, the improvements arising from the relationship at least have the potential to enhance the competitiveness of the entire district.
CONCLUSION This chapter describes an organizational innovation in which an SME is established as a key supplier to a leader firm. It shows how the development of the relationship between the two firms resulted in the success and growth of the SME. This is, at its simplest, stark proof of the possibility of development of indigenous, start-up SMEs even in the context of globalization, and even in an economy like Ireland’s, dominated by multinational corporations. Aspects of the case are unusual. Differing markedly from some of the leader firm–supplier relationships of the Italian industrial districts, in particular the Benetton oligopsony where links with second and third tier suppliers are subject to unilateral controls, Oceanpath and Superquinn enjoy a collaborative, reciprocal and bilateral arrangement. Between the leader firm and its supplier a contextual environment has been created that has transformed a purely transactional (supply chain) relationship into a collaborative, knowledge sharing and ultimately innovation inducing one. The success of the relationship can be attributed to bilateral trust, openness and commitment. There is a strong temporal dimension however to the creation of this particular contextual setting. Over time, interpersonal trust evolved into process-based trust. This atmosphere of
198
SMEs in a globalised world
trust has ensured the successful coordination of the relationship without contract. Furthermore, a mutually beneficial partnership has been created facilitating improvements in the supplier’s strategic and managerial functions (as in the flagship model) while at the same time bridging the gaps in technical information via the leader firm’s external sources of knowledge (Boschma and Lambooy, 2002). Acting as a source of external knowledge or a channel through which external knowledge is received the leader firm has led the supplier to innovate in a number of different ways, for example, automating, traceability and new product development (value added products). In stark contrast to a monopsony where the power resides in the leader firm, Oceanpath has been able, in the words of its business relations manager, to ‘cash in’ on their association with the retailer and win new custom from other retail multiples. In return, the leader firm enjoys the benefits of vertical integration without vertically integrating – what may be termed quasi-vertical integration (Langlois and Roberston, 1995). Without the relationship with Superquinn it would not have been possible for Oceanpath, a small firm, to have developed and grown in the way it has. We can conclude therefore, that in this case, the leader firm is a ‘good’ one. In reaching this conclusion however we must add a cautionary note outlining that not all leader firm–supplier relationships are necessarily beneficial to small firms and their development particularly where the relationship is bureaucratic in its construction and where scope for relational contracting is negligible.
NOTES 1. Among early writers on the threats to the canonical form was Harrison (1994a; 1994b). 2. In real terms. 3. Coordination integration is the term used by Langlois and Roberston (1995) to describe integration other than ownership. It refers to various cooperative exchanges among autonomous economic actors. 4. The interested reader should refer to Grandori (1997) for further detail. 5. In the Irish context a small firm is one that employs between 10 and 49 (Morgenroth and O’Malley 2003). 6. Contextual environment is institutional, comprising social institutions, conventions, norms and trust – non-physical features that are embedded in place and specific to a given location. The contextual environment shapes and frames the environment within which transactional and non-transactional inter-firm relationships occur (McGrath, 2006, p. 219). 7. BIM is Bord Iascaigh Mhara – The Irish Fisheries Board. 8. Musgraves is a wholesale supplier to, and owner of, chains of mini-markets. 9. For more on the industrial district type characteristics of fish processing in north Dublin, see Jacobson and McGrath (2006).
A case study of a leader firm–supplier relationship
199
REFERENCES Alberti, F.G. (2006), ‘The decline of the industrial district of Como: recession, relocation or reconversion?’, Entrepreneurship and Regional Development, 18 (6), 473–501. Albino, V., A.C. Garavelli and G. Schiuma (1999), ‘Knowledge transfer and inter-firm relationships in industrial districts: the role of the leader firm’, Technovation, 19 (1), 53–64. Becattini, G. (1991), ‘Italian industrial districts: problems and perspectives’, International Studies of Management and Organization, 21 (1), 83–91. Bender, G. and S. Laestadius (2005), ‘Non-science based innovativeness: on capabilities relevant to generate profitable novelty’, paper presented at the PILOT conference Low-Tech as Misnomer: The Role of Non-Research-Intensive Industries in the Knowledge Economy 29–30 June, Brussels. Boschma, R.A. and J.G. Lambooy (2002), ‘Knowledge, market structure, and economic co-ordination: dynamics of industrial districts’, Growth and Change, 33 (Summer), 291–311. Brusco, S. (1982), ‘The Emilian model: productive decentralisation and social integration’, Cambridge Journal of Economics, 6 (2), 167–85. Camuffo, A. (2003), ‘Transforming industrial districts: large firms and small business networks in the Italian eyewear industry’, Industry and Innovation, 10 (4), 377–401. Capello, R. (2002), ‘Spatial and sectoral characteristics of relational capital in innovation activity’, European Planning Studies, 10 (2), 177–200. Carbonara, N. (2002), ‘New models of inter-firm networks within industrial districts’, Entrepreneurship and Regional Development, 14 (3), 229–246. Carbonara, N. (2004), ‘Innovation processes within geographical clusters: a cognitive approach’, Technovation, 24 (1), 17–28. Coro, G. and R. Grandinetti (2001), ‘Industrial district responses to the network economy: vertical integration versus pluralist global exploration’, Human Systems Management, 20 (2), 189–99. Dei Ottati, G. (1991), ‘The economic bases of diffuse industrialization’, International Studies of Management and Organisation, 21 (1), 53–74. Dwivedi, M., R. Varman and K.K. Saxena (2003), ‘Nature of trust in small firm clusters’, International Journal of Organizational Analysis, 11 (2), 93–104. Garibaldo, F. and D. Jacobson (2005), ‘The role of company and social networks in low tech industries’, Perspectives on Political, Social and Economic Integration, 11 (1–2), 233–70. Grandori, A. (1997), ‘An organizational assessment of interfirm coordination modes’, Organization Studies, 18 (6), 897–925. Harrison, B. (1994a), ‘The Italian industrial districts and the crisis of the cooperative form: part I’, European Planning Studies, 2 (1), 3–23. Harrison, B. (1994b), ‘The Italian industrial districts and the crisis of the cooperative form: part II’, European Planning Studies, 2 (2), 159–75. Heanue, K. (2007), ‘Industrial agglomeration in the Irish furniture industry’, unpublished manuscript, Rural Economy Research Centre, Teagasc, Athenry, Ireland. Heanue, K.P. and D. Jacobson (2008), ‘Embeddedness and innovation in low and medium technology rural enterprises’, Irish Geography, 41 (1), 113–37.
200
SMEs in a globalised world
Jacobson, D. and H. McGrath (2006), ‘Localising economic development in North Dublin: lessons from three industrial sectors’, in D. Jacobson, P. Kirby and D. O’Broin (eds), Taming the Tiger: Social Exclusion in a Globalised Ireland, Dublin: tasc and New Island, pp. 139–53. Jacobson, D. and Z. Mottiar (1999), ‘Globalisation and modes of interaction in two sub-sectors in Ireland’, European Planning Studies, 7 (4), 229–45. Jacobson, D. and D. O’Sullivan (1994), ‘Analysing an industry in change: the Irish software manual printing industry’, New Technology, Work and Employment, 9 (2), 103–14. Kelly, K. (2002), ‘Carolan the collaborator’, Checkout, November, accessed 6 April 2008 at www.checkout.ie/Feature.asp?ID = 69. Koschatsky, K. (1999), ‘Innovation networks of industry and business related services – relations between innovation intensity of firms and regional inter-firm cooperation’, European Planning Studies, 7 (6), 737–57. Langlois, R.N. and P.L. Robertson (1995), Firms, Markets and Economic Change: A Dynamic Theory of Business Institutions, London: Routledge. Lazerson, M.H. and G. Lorenzoni (1999), ‘The firms that feed industrial districts: a return to the Italian source’, Industrial and Corporate Change, 8 (2), 235–67. Lorenzen, M. (2002), ‘Preface’, special issue on Clustering, Capabilities and Coordination in Industrial Clusters, International Studies of Management and Organization, 31 (4), 3–11. Maskell, P. and A. Malmberg (1996), ‘The competitiveness of firms and regions: ubiquitification and the importance of localised learning’, European Urban and Regional Studies, 6 (1), 9–25. McGrath, H. (2006), ‘Industrial clusters in local and regional economies: a post-Porter approach to the identification and evaluation of clusters in North Dublin’, unpublished PhD, Dublin City University. Morgenroth, E. and E. O’Malley (2003), ‘SMEs and regional development in Ireland’, in B. Fingleton, A. Eraydin and R. Paci (eds), Regional Economic Growth, SMEs and the Wider Europe, Aldershot: Ashgate Publishing, pp. 161–91. Paniccia, I. (1998), ‘One, a hundred, thousands of industrial districts: organisational variety in local networks of small- and medium-sized enterprises’, Organization Studies, 19 (4), 667–99. Piore, M. and C. Sabel (1984), The Second Industrial Divide: Possibilities for Prosperity, New York: Basic Books. Rugman, A.M. and J.R. D’Cruz (2000), ‘The theory of the flagship firm’, in D. Faulkner and M. de Rond (eds), Cooperative Strategy: Economic, Business and Organizational Issues, Oxford: Oxford University Press, pp. 57–73. Russo, M. (2004), ‘The ceramic industrial district facing the challenge from China’, paper prepared as a contribution to the research project ‘Distretti industriali come sistemi complessi’ [‘Industrial Districts as Complex Systems’], accessed 6 April 2008 at www.economia.unimore.it/convegni_seminari/CG_sept03/Papers/ Parallel%20Session%201.1-2.1-3.1/Russo.pdf. Sabel, C.F. (2004), ‘Districts on the move: notes on the Tedis Survey on the internationalization of district firms’, paper presented at the ‘Local Production and Governance Conference’, December, Turin. Sammarra, A. and F. Belussi (2006), ‘Evolution and relocation in fashion-led Italian districts: evidence from two case-studies’, Entrepreneurship and Regional Development, 18 (6), 543–62.
A case study of a leader firm–supplier relationship
201
Schmitz, H. (2000), ‘Does local co-operation matter? Evidence from industrial clusters in South Asia and Latin America’, Oxford Development Studies, 28 (3), 323–36. Sengenberger, W. and F. Pyke (1992), ‘Industrial districts and local economic regeneration: research and policy issues’, in W. Sengenberger and F. Pyke (eds), Industrial Districts and Local Economic Regeneration, Geneva: International Institute for Labour Studies, pp. 3–29. Williamson, O.E. (1985), The Economic Institutions of Capitalism: Firms, Markets and Relational Contracting, New York: Free Press.
10.
Dynamics of the SME sector in Ireland: a driver of growth in the Irish economy since 1994? Helena Lenihan, Briga Hynes and Mark Hart
INTRODUCTION The extraordinary growth trajectory of the Irish economy (the ‘Celtic Tiger’) since 1994 has attracted a great deal of interest, commentary and research both within and outside of Ireland. The Central Bank of Ireland reports for 1992 and 1999 clearly highlight the scale of this transformation. In the 1992 Annual Report it states ‘Real GNP growth seems likely to be in the region of 2 per cent in 1993 slightly higher than the 1.75 per cent estimated for 1992 . . . The standardised Live Register unemployment rate will probably average 17.75 per cent for 1993 compared to 16 per cent in 1992’ (Central Bank of Ireland, 1992, p. 9). By the summer of 1999, the assessment had changed dramatically ‘The Irish economy continues to expand at a very rapid pace . . . the volume of GNP is estimated to have grown by about 8 per cent last year and further growth of about 6.5 per cent is projected for 1999, with a similar rate of growth likely next year’ (Central Bank of Ireland, 1999, p. 5). Irish GNP per head in 1987 stood at 59 per cent of the EU15 average, largely unchanged from its 1960 position, but by 1997 it had risen to 88 per cent (Barry, 1999, p. 1). This growth was clearly transmitted to the labour market as over the same period the numbers at work increased at an annual average rate of 2.1 per cent, and the unemployment rate fell from 17.1 to 10.3 per cent (Barry, 1999; Harris, 2005). It was perhaps not surprising that many countries looked to Ireland as an economic development role model, and the Sapir Report (Sapir et al., 2003) suggested that Ireland could provide key lessons for other EU countries with regards to realising the objectives set out in the Lisbon Agenda. In a similar vein, Andreosso-O’Callaghan and Lenihan (2006) and Bailey et al. (2007) have debated whether the industrial policy approach adopted in Ireland could be viewed as providing lessons for some of the newer EU member states (for example, Hungary, Slovenia and Slovakia). 202
Dynamics of the SME sector in Ireland
203
However, much of the discussion of this late-twentieth century economic growth ‘miracle’ in Ireland, has, from an industrial policy perspective, been on the importance of foreign direct investment (FDI) inflows (for example, Gray, 1997; Barry, 1999 and Ruane and Görg, 1999; World Bank Group, 1999; Harris 2005; Acs et al., 2007) and to a lesser extent on the performance of an indigenous stock of small firms. According to the Small Business Forum Report (2006), there were approximately 250 000 small firms in Ireland (inclusive of both manufacturing and service firms) in 2005. Furthermore these firms accounted for more than 97 per cent of all enterprises in Ireland. Given the numerical importance of SMEs in Ireland a lack of discussion of their role in the rapid growth of the economy is surprising, particularly in light of the extensive debate that has taken place internationally in academic and policy circles about the relationship between entrepreneurship and economic development (Acs and Storey, 2004; Fritsch and Mueller, 2004; Plummer and Acs, 2004; Thurik and Wennekers, 2004; and van Stel and Storey, 2004). In this chapter we refer to unpublished annual Irish VAT data for the period 1988–2005 to provide a detailed look at national and regional trends in SME business birth and death rates in Ireland. We are able, therefore, to address a hitherto overlooked research question in Ireland: the extent to which SME business birth and death rates are connected to the rapid growth and expansion of the Irish economy. Furthermore, it is important that the academic and policymaking communities examine the dynamics of the SME sector in Ireland in the context of a corresponding increase in Ireland’s cost base (Forfás, 2006). Ireland in terms of its perception as a competitive location for FDI is facing increasing competition in attracting inward FDI from low cost production countries such as Eastern Europe and Asia, again reinforcing the need to have an established small firm sector operating in tandem with the larger FDI firms. While globalisation and international market liberalisation have opened up many markets to Irish business, they have also resulted in a greater international presence in Ireland posing new challenges that will have an influence on the sustainability of the SME sector in the years ahead. In summary therefore, the overriding objective of this chapter is to establish whether the SME sector has played a role in the economic transformation of the Irish economy since 1994. The chapter is structured as follows. The first section provides a brief profile of SME activity in Ireland over the period 1994–2005 in terms of their relative share of the total business stock. We also examine the sectors in which SMEs operate and analyse the contribution of SMEs to export performance. This helps to establish the extent to which SMEs have indeed been an important contributory driver of growth in the
204
SMEs in a globalised world
Irish economy since 1994. The second section examines the relationship between entrepreneurship and economic development in Ireland using unpublished annual Irish VAT data for the period 1988–2005 to provide a detailed look at national and regional trends in SME business birth and death rates in Ireland. The third section explores the policy arena in which SMEs operate in Ireland and we outline the various industrial policy interventions that have been aimed at the SME sector. The fourth section addresses the issue of the effectiveness of such interventions. This evaluation theme is important given the increased impetus from the EU for national governments to engage in policy evaluation to ensure accountability and also to address the fundamental issues of value for money and opportunity cost. The chapter concludes with a discussion of how such policy interventions address the needs of the SME sector in Ireland to ensure small firms are encouraged to grow. This is particularly the case for firms operating in high growth value added industry sectors and those small businesses that can be categorised as high potential start-ups (HPSUs) and those with a focus on international markets. Finally, we assess in more detail what the Irish case tells us about the wider international research agenda linking entrepreneurship and economic growth.
A PROFILE OF SME ACTIVITY IN IRELAND: 1994–2005 The importance of the established SME sector to the Irish economy has been acknowledged both in the research of Gudgin et al. (1995), Barkham et al. (1996) and Kinsella et al. (1994) and from a policy perspective in The Small Business Forum Report (2006). Despite this, there is a still a lack of comprehensive data profiling the wider contribution of the small firm to the Irish economy apart from data on their contribution to employment. The following discussion provides for the first time an overview of the contribution of the SME sector extending beyond its employment contribution to profile trends in the growth in the number of SMEs and an examination of SME participation in export activity during the period. The Task Force on Small Business (Government of Ireland, 1994b) produced the first in-depth report on the profile and the contribution of the SME sector in Ireland at that time. This study estimated there were 160 000 non-farm small businesses operating in Ireland in 1994, which employed a little over one-third of the workforce (435 000 people). Furthermore, 98 per cent of these firms employed less than 50 persons, and in excess of 85 per cent of firms employed fewer than 10 people. The majority of these
Dynamics of the SME sector in Ireland
205
firms (80 per cent) operated in the service sector at that time (Government of Ireland, 1994b, p. 12). Data produced by the Revenue Commissioners (2004) indicated that in the year 2002, there were just over 283 000 small enterprises in Ireland, comprising approximately 243 000 sole traders and partnerships, and 40 000 limited companies. By comparison, estimates provided by the Central Statistics Office (CSO) (2006b) suggested that ten years on from the publication of the ‘Task Force on Small Business Report’ that the number of small firms had increased to approximately 250 000 accounting for 97 per cent of all enterprises in Ireland. These firms employed 777 000 people, or 54 per cent of the total private sector workforce in Ireland. This suggested an increase in their contribution to employment of 79 per cent over the ten years. On an aggregate basis these trends indicate that the number of small firms had grown by over 50 per cent between 1994 and 2005 (from 160 000 to 250 000 SMEs). In contrast to the manufacturing sector which remained virtually static in terms of business stock, the service sector experienced substantial growth in Ireland over the last decade (Small Business Forum, 2006). Data on SME activity in this sector is only available from the year 2000 onwards and is obtained from the Annual Services Inquiry (CSO, 2006a) surveys. In the year 2000 there were approximately 61 700 small service firms in operation (employing between 1–49 persons). The number increased to 81 506 firms in 2003 and to 83 498 firms in 2004. The greatest increase in the number of firms in this sector was over the period 2000 to 2003 (approximately 34 per cent increase). In 2004 there were 83 498 service firms in operation, reflecting an increase of only 1992 firms compared with previous years. With regard to their contribution to employment in the year 2000, 290 600 persons were engaged in service firms with less than 50 persons representing 52.5 per cent of overall employment in the sector. In 2003, the number of persons employed in the service sector small firms was 384 600 which represented 54 per cent of overall employment, reflecting a very slight increase in the share from the previous year. Total employment in this sector increased by over 183 000 in the period 2000 to 2004, where more than half of the total workforce (473 000) of the services sector worked in small firms in 2004. This was an increase in employment of 112 000 persons since the year 2000. In summary, it is clear that the growth in the number of SMEs in Ireland was stronger in the 1995–2000 period, displaying approximately a 6.6 per cent per annum increase in the number of SMEs. However, of note was the decline in the growth in the number of SMEs in operation in 2000–2005 (a growth rate of 2.6 per cent per annum). Further investigation revealed that less positive growth was evident in the manufacturing
206
SMEs in a globalised world
sector. In comparison the service sector had the highest concentration of small enterprises, where the vast majority of firms (98 per cent) employed less than 20 people. In addition to growth in the service sector was a corresponding growth in SME activity in the construction sector, which accounted for a quarter of all small businesses in operation and 27 per cent of those employed in small firms in 2005. Finally, the analysis of SME activity highlighted the significance of the micro firm (less than 10 employees) as a central component of the overall SME sector in Ireland. A characteristic of the business environment in Ireland between 1994 and 2005 was an increasing emphasis on the need for Irish firms irrespective of size to consider international markets as a means of firm growth (Lawless, 2007; Enterprise Strategy Group, 2004).
BUSINESS START-UPS AND ECONOMIC GROWTH IN IRELAND Perhaps surprisingly, what has been almost completely missing from industrial policy discussions on the ‘Celtic tiger’ is a consideration of new business ventures and entrepreneurship (besides the recent work of Burnham, 2003; Ferreira and Vanhoudt, 2004; and Acs et al., 2007, whose work provides only limited insights into the key research question discussed below). In the context of other economies, this discussion has taken place (see for example, the work of Wennekers and Thurik, 1999; Audretsch and Thurik, 2001; Acs and Storey, 2004; Thurik and Wennekers, 2004;) but this has largely not been the case for Ireland. Recent work has attempted to examine the issue of business formation and survival rates in the case of the rapidly growing Irish economy (Anyadike-Danes et al., 2010). The analysis is based on a regional analysis of Irish Value Added Tax (VAT) data over the period 1988 to 2005 (the annual data was provided by the Irish Revenue Commissioners). This data provides a unique opportunity to understand more clearly the significance of new venture creation to past and emerging spatial trends in an Irish context. The key research question addressed was as follows: are big business birth rates, where they arise, a cause of economic progress in those places or an effect (that is, a by-product of the factors driving the growth process)? Anayadike-Danes et al. (2010) use aggregated annual data on business ‘births’ and ‘deaths’ (as proxied by registrations, re-registrations and cancellations) to address the above research question. Given the phenomenal growth of the Irish economy 1994–2000, one would expect to find some evidence that business birth rates are in some manner positively connected to that trend. At the start of 1988, there were just over 75 000 VAT-
Dynamics of the SME sector in Ireland
207
registered businesses in Ireland and by the end of 2004, there were around 220 000 (the stock of businesses had about tripled in less than 15 years). Moreover, the growth in the business stock surpassed population growth by a considerable degree. As outlined by Anyadike-Danes et al. (2010) in 1988 there were 22 businesses per thousand population, and by the end of 2004, there were 56 businesses per thousand. What is interesting is that at the national level, the growth rate in new business venture creation in Ireland in the late 1980s and early 1990s (that is, pre-‘Celtic tiger’ period) is broadly in line with those seen in later periods (that is, either the rapid growth phase period 1994 until 2000 or the period 2001 until 2004). More precisely, given the investigation of the link between new venture creation and rapid economic growth in Ireland, the vital rates (gross and net births and deaths) record similar rates of growth at the beginning (1988), the middle (1998) and the end (2004) of the period (Anyadike-Danes et al., 2010). This demonstrates that economic growth can occur at the national level without a concomitant increase in business start-up activity. Whilst the stock of businesses in Ireland has grown in number over the period 1988 to 2004, the raw numbers themselves are a little misleading since after standardisation a continuation of the growth rates experienced in the late 1980s would have produced a similar sized stock of businesses as was actually recorded in 2004. Therefore, there does not appear to be any evidence of a ‘Celtic tiger’ effect in business stock growth. By this we mean that the observed dynamic of the population of businesses in Ireland over this period would appear to be operating independently of factors which were driving Ireland’s rapid economic growth. From this we conclude that Ireland’s economic growth was not driven by a rise in new venture creation, nor can we associate this period of rapid economic growth as providing a stimulus to increased gross or net birth rates. Lawless and Murphy (2008) provide some supporting evidence for this conclusion when they state that higher average rates of job creation in the manufacturing sector between 1972 and 2006 are not solely associated with periods of high national employment growth. This is an important conclusion for policymakers as they seek to develop enterprise policies designed to stimulate economic growth through the process of business start-up interventions.
THE INDUSTRIAL POLICY ARENA IN WHICH IRISH SMES OPERATE Discussion here will primarily concern itself with the industrial policy arena in which SMEs in Ireland have operated in since the mid-1990s. The
208
SMEs in a globalised world
key question to investigate is the extent to which industrial policy can be positively associated with the growth in the number, nature and size of SMEs. The discussion begins by briefly reviewing the various stages of the Irish industrial policy approaches. According to Drudy (1995), three broad approaches have been adopted in the Irish economy; 1. 2. 3.
the ‘protectionist’ era roughly covers the period from the 1930s to the mid-1950s; the ‘outward looking’ era covers from the 1950s and extends onwards; an ‘enterprise and market-oriented’ approach also became obvious from the early-1980s.
Clearly these are not mutually exclusive categories with clear cut-off dates. They do, however, provide a useful framework with which to review the various industrial policy changes and developments. From the early-1950s, there was a realisation of the difficulties and drawbacks associated with relying on protectionist and self-sustaining policies which characterised the post-independence decade. While protectionism established an industrial base, it remained a weak and vulnerable one. Building on the report of Whitaker (1958) the First Programme for Economic Expansion was launched in 1958. This resulted in a more outward looking industrial policy approach in an attempt to develop a successful industrial sector. With tax breaks and subsidies, the Irish government encouraged foreign companies to set up branches in Ireland. In 1973, the Industrial Development Authority (IDA) concentrated on attracting electronics, chemicals and other ‘high-technology’ sectors. From the 1980s several reports have been written about Irish industrial policy. The Telesis Consultancy Group (1982) cited criticisms such as: little discussion of regional policy and over-emphasis on capital grants and foreign industry. The main recommendations of the Telesis report were given support in the White Paper on Industry Development (Government of Ireland, 1984). The Culliton report (1992) called for a ‘holistic’ approach to economic development by stressing the optimisation of factors that drive firm competitiveness. These included for example, reform of taxation and education systems and infrastructural improvements. On a local basis, the establishment of City and County Enterprise Boards, as well as Enterprise Areas, in a number of urban areas provided solid evidence of the new approach. The Industrial Development Act of 1993 restructured the promotion agencies; ‘Forbairt’ was established to promote indigenous industry and the IDA was given a mandate to attract foreign direct investment. As outlined
Dynamics of the SME sector in Ireland
209
by Andreosso-O’Callaghan and Lenihan (2006), ‘Culliton placed emphasis on the creation of an overall competitive business environment with the hope that firms would move away from the grant dependency mentality identified by previous reports’ (p. 279). Also significantly (and particularly in light of the focus of this chapter on SMEs), Culliton also emphasised the need for the expansion of the indigenous sector. Other industrial policy statements came later in the form of the various National Development Plans (1989–1993 and 1994–1999) and Operational Programmes for Industry (co-financed by the EC). In 1993, the ‘Report of the Task Force on Jobs in Services’ (Government of Ireland, 1994c) emphasised a new policy was needed for the oft neglected services sector. The ‘Task Force on Small Business’ (Government of Ireland, 1994b) represented the first report whose sole focus was on small business. This was a significant development in that it represented the first formal recognition by the Irish government about the importance of SMEs as a key contributor to growth and employment within the Irish economy. This is astonishing when one considers that as far back as 1979 some 95 per cent of all manufacturing units in Ireland could be classified as SMEs (Andreosso-O’Callaghan and Lenihan, 2006). This overlapped with an increasing academic awareness (for example, Capecchi, 1990) in Europe with respect to the gradual shifting of economies to what can be described as a post-Fordist mode of production organisation which is characterised by factors such as inter alia: small firms; small batches of production as opposed to increasing returns to scale and flexible specialisation and innovation strategies (normally associated with small firms). In 1995, the ‘Small Business Operational Programme’ (Government of Ireland, 1995) was also published. The report Shaping our Future – A Strategy for Enterprise in Ireland in the 21st Century (Forfás, 1996) provided a framework for the development and promotion of an enterprise sector, a project that heretofore had never been the focus of industrial policy in Ireland. The ‘Operational Programme for Industrial Development 1994–1999’ (Government of Ireland, 1994d) focused on helping firms, particularly, small, indigenous firms, increase sales over the five year period 1994–1999. In July 1998, Forbairt ceased to exist and a new agency with responsibility for indigenous industry, Enterprise Ireland, was established. More recently, the National Development Plan 2000–2006 (Government of Ireland, 2000) proposed spending more on training and personal development than on job creation grants to industry, agriculture, tourism, forestry and fisheries combined. In July 2003, the Irish government formed a strategy group with the key remit of examining ways of strengthening Irish competitiveness by promoting a knowledge-driven economy. The overriding recommendation to emanate from the Group’s report (Enterprise
210
SMEs in a globalised world
Strategy Group, 2004) was that business issues should be placed at the core of government policymaking. This recommendation was echoed in the findings of the Small Business Forum (2006) Report. This report was the output of the Small Business Forum, established by the Irish government in 2005 with the objective of evaluating the current business environment for the small firm and to advise on the appropriateness of public policy initiatives including the interventions by the enterprise development agencies. Essentially the findings in this report reinforced the importance of the SME sector as a key component of the industrial base in Ireland. Furthermore, the report inter alia endorsed the need to encourage more innovative knowledge-based firms with the ability to compete in an increasingly competitive international market. The above has provided a broad (macro) overview of the various industrial policy statements (some of which are more pertinent than others in terms of their support/focus on SMEs) with regards to the Irish economy. In the discussion that follows, we now try to look at the more micro interventions (many of which are contained in, or have emanated from the various industrial policy statements referred to above) that have taken place with respect to the Irish industrial policy playing field and more specifically, those directed at SMEs. The monetary extent of public policy interventions with respect to industrial development in Ireland is clearly visible when one considers that as outlined by Lenihan et al. (2005), ‘It has been estimated that approximately €5.5 billion was spent on direct financial assistance to industry by government agencies (excluding administration and support to the higher education sector) over the period 1980–2003’ (p. 70). A feature of the recent industrial policy intervention landscape has been the significant private equity investment (including seed and venture capital funding) that has taken place during the 1980s and early-1990s. As outlined by Lenihan et al. (2005) ‘The latter 1990s witnessed a substantial increase in private equity investment in Ireland, with payments increasing from €17.8 million in 1996 to €32.8 million in 1997’ (p. 71). Some of the more recent industrial policy documents continue to emphasise the relevance of financial assistance to industry. The focus is increasingly somewhat different from that which went before it with increased emphasis nowadays on ‘capability building’ and support for applied R&D as opposed to ‘capacity expansion’. In terms of the rationale for government intervention (largely with respect to financial assistance to firms), the justification is usually made with respect to the existence of market failure (more specifically the existence of incomplete markets).1 It has been widely argued that market failure exists in the conventional markets for small firm finance for example, thus
Dynamics of the SME sector in Ireland
211
providing an underlying rationale for government intervention in such markets (Storey, 1994). Incomplete markets have also been much discussed in terms of a gap in the availability of private sector equity finance (formal or informal) to firms (in many cases particularly small firms) for start-up and expansion activity (see Mason and Harrison, 1992; Mason and Harrison, 1997; Harding, 2000 amongst others). Indeed, several studies on the barriers encountered by small firms in Ireland have pointed to access to finance as being the single most critical issue (Fitzgerald and Breathnach, 1994; Goodbody Economic Consultants, 2002; Small Business Forum, 2006) As outlined by Hart and Lenihan (2006), ‘One of the key objectives of industrial policy, in Ireland as well as elsewhere, is to close that gap by the provision of equity finance to the private sector and in particular SMEs’ (p. 334). More specifically, the Irish government has tried to address the problem of incomplete markets for equity finance for Irish firms with a variety of initiatives in the past 20 years or so, beginning with the Government ‘White Paper on Industry Development’ (Government of Ireland, 1984) which proposed a method of direct state equity in new or existing companies through the National Development Corporation. The year 1984 also witnessed the establishment of the Business Expansion Scheme to encourage equity investments by individual taxpayers. Throughout the 1980s lively debate was evidenced in various government policy documents (‘Community Support Framework for Industrial Development 1989–1993’ (Commission of the European Communities, 1989); ‘National Development Plan 1989–1993’ (Government of Ireland, 1989)) regarding the equity gap pertaining to small firms. The influential Culliton Report (1992) also played a strong part in recommending a shift away from grants to indigenous industry in favour of an expansion of equity and venture capital activities by state agencies. Shortly afterwards, the Industrial Credit Company (ICC) bank established a subsidiary company ICC Venture Capital with the prime concern of managing a number of specialist venture capital funds. The 1993 Finance Act also introduced the Seed Capital Scheme. Given the time period under investigation in the current chapter, it is indeed interesting to note that the Operational Programme for Industrial Development 1994–1999 (Government of Ireland, 1994) also called for greater use of equity (including preference shares in the provision of support to industry). Interestingly, the report outlined that the lack of growth pertaining to a number of SMEs in Irish manufacturing was the result of a lack of equity capital. Even more recently, the ‘National Development Plan 2000–2006’ (Government of Ireland, 2000) witnessed the introduction of the Seed and Venture Capital Programme. The Enterprise Strategy Group (2004) once again recommended that state
212
SMEs in a globalised world
intervention move away from direct support for second round financing and leave that part of the market to private sector venture capital investors. A report from the Small Business Forum in 2006 highlighted that owner/managers still reported problems in obtaining suitable finance for growth. This problem was most pronounced in the service sector and for SMEs who were not eligible for state supports. The Forum recommended extension of the Business Expansion Scheme. With regards to the Seed Expansion Scheme, it recommended the development of a network of ‘Business Angels’ comprising of private investors to bridge this finance gap.
EVALUATING THE EFFECTIVENESS OF SME POLICY INTERVENTIONS IN IRELAND Surprisingly, despite the vast sums of money invested in the industrial sector in Ireland through government schemes to support start-up and growth, there has been a dearth of research which has concerned itself with measuring the impact of such interventions. It is beyond the focus of the current chapter to discuss in detail the reasons why Ireland has such a poor evaluation culture but it is worth noting that it may relate to issues such as the characteristics of the Irish economics research community (applied microeconomic analysis – specifically industrial microeconomic analysis not particularly well developed; small size of the country; weak history of planning; inclusive negotiated agreements; political tradition of client focus (for a more in-depth discussion, refer to Lenihan et al., 2005 and Ruane, 2004). To begin, we review some of the studies which have concerned themselves with evaluating the impact of Irish industrial policy interventions. The discussion will concern itself primarily with the examination of the evaluation of SME policy interventions and in particular those that have taken place since 1994. Defining Evaluation It is beyond the remit of the current chapter to delve into the debate as to the definition of ‘evaluation’ (for a more in-depth discussion on this issue, the interested reader should refer to Storey, 2000). For the purpose of this chapter, we adopt the approach set out by Lenihan et al. (2005) where evaluation is described as ‘the process of assessing the level of “additionality” associated with a particular policy intervention, and our primary focus will be at the micro or firm level’ (Lenihan et al., 2005, p. 69). The
Dynamics of the SME sector in Ireland
213
authors proceed to argue that the thrust of the term additionality is that a policy intervention by government should bring about a level of economic activity (for example, improved innovation or productivity or increased numbers employed) in excess of what would have happened anyway in the absence of the specific intervention – the counterfactual position. Given the definition of ‘evaluation’ adopted for the purpose of this chapter, our particular focus here is concerned with those studies that measure the level of ‘additionality’ associated with SME interventions. More specifically, we are interested in examining SMEs in the post-1994 period. The increased impetus for evaluation during the 1990s was primarily driven by the EU who emphasised the need to assess accountability and value for money so as to measure the impact of the significant EU transfers received by Ireland. The 1990s also witnessed the establishment of the Industrial Evaluation Unit (IEU) in Ireland and also the National Development Plan/Community Support Framework (NDP/ CSF) Evaluation Unit. A central element of many of the Irish evaluation studies to date has been the assessment of what would have happened in the absence of a particular type of public policy intervention (mostly in the form of financial assistance to firms). The key objective is thus to explore and develop a counter-factual scenario. The exploration of the counter-factual embodies the concepts of deadweight and displacement. As outlined by Lenihan et al. (2005) ‘Jointly, these concepts allow an assessment of the additional impact of any policy intervention’ (p. 73).2 Meaning of the Terms Deadweight and Displacement The term displacement is widely accepted as implying the degree to which output from an assisted firm displaces output from already existing firms in the marketplace. On the other hand, ‘deadweight’ as a concept has far less certainty surrounding it.3 In line with the work of Lenihan (1999; 2004), we take deadweight to mean the degree to which economic activity (for example, increased sales or productivity) at the level of the firm would have happened in the absence of intervention by public agencies. Table 10.1 summarises the findings of a selection of what we see as some of the most significant evaluation studies that have been undertaken since the mid-1990s. As outlined previously, discussion here only pertains to those findings which concern themselves with interventions to support SMEs. From Table 10.1 below, it is clearly evident that deadweight estimates in the Irish economy are high. Lenihan (2004) argues that they are indeed high when compared with those derived in international studies, where evidence suggests that estimates range from 8 per cent to 59 per cent (see
214
Table 10.1
SMEs in a globalised world
Estimates from Irish evaluation studies which have derived or incorporated estimates of deadweight as applied to the case of SMEs
Authors (year)
Focus of the evaluation study
Deadweight estimate
Industrial Evaluation Unit (1999)
Micro enterprise supports across national and local development agencies
45%
Industrial Evaluation Unit (2000)
Seed and Venture capital scheme
60%
Fitzpatrick Associates Economic Consultants (2004)
Role of the County and City Enterprise Boards (study looked at inter alia the net impact of supports for micro enterprise)
Various estimates of deadweight ranging from figures of 30% for pure deadweight and 60% for partial deadweight to other indirect measures with the study arriving at a pure deadweight factor of approx 27% and a partial deadweight factor of about 41%.
Lenihan and Hart (2004)
Estimate the net additionality of financial assistance from Enterprise Ireland to indigenously-owned firms in Ireland for the period 2000–2002.
46.2%–55.8% Many of the firms in the sample were High Potential Startups (HPSU) with mean employees of 22.
Lenihan (1999)
Estimate of deadweight in firms that received grants from Shannon Development in 1995
78.4% Indigenous Irish firms (91% of which are SMEs).
Lenihan 2004 for detailed discussion).4 Hart and Lenihan (2006) in their research on the venture capital industry in Ireland have also emphasised the importance of accounting for positive externalities (for example, leverage of additional equity finance, benefits of having the backing of a state development agency) when estimating additionality and deadweight. In fact, evidence from their work on Enterprise Ireland assisted high potential
Dynamics of the SME sector in Ireland
215
start-ups (HPSUs) over the 2000–2002 period has shown that accounting for positive externalities can indeed ‘boost’ the estimate of additionality between 5 and 7 percentage points. They also report that it was ‘small, young high-tech firms that were more likely to report that Enterprise Ireland financial assistance resulted in the leverage of additional equity finance’ (Hart and Lenihan 2006, p. 343). In light of such high deadweight estimates related to Irish industrial policy interventions, it is interesting to know what the underlying causes are. Lenihan (2004) and Lenihan and Hart (2006) have employed Logit regression analysis to estimate predictive models of deadweight and displacement based on firms that received grants from Shannon Development in 1995. These studies find that size of firm impacts on deadweight. More precisely, firms with 0–49 employees (small firms) are less likely to influence deadweight than firms with greater than 50 employees (larger firms). In the case of displacement, the studies find that size is also a significant variable. More specifically, the findings suggest that larger firms (that is, greater than 50 employees) are more likely to cause displacement ceteris paribus. Lenihan (1999) estimated that in the case of indigenous firms (91 per cent of which were classified as SMEs) grant aided by Shannon Development in 1995, the displacement estimate was 4.2 per cent. More recently, a study by Fitzpatrick Associates Economic Consultants (2004) which reviewed the role of County Enterprise and City Enterprise Boards (and which raised concerns about the net impact for micro enterprise supports) estimated that in the case of displacement, 57 per cent of turnover emanated from the home country or the promoter, with a further 33 per cent emanating from the rest of Ireland. Thus, they conclude by saying that the results of their study suggest a potentially higher level of displacement to that estimated by Lenihan (1999). More generally, Hart and Lenihan (2006) argue that undoubtedly, ‘the development of a seed and venture capital fund industry in Ireland (aided by EU funding) has been one of the key improvements in term of financing Irish industry’ (p. 337). They report that Enterprise Ireland’s (HPSUs) client firms, which were more likely to be smaller and involved in high-tech activities, were also more likely than more established firms to use equity finance and most especially formal venture capital to finance their projects (59 per cent of total project costs for HPSUs versus 11.1 per cent for expansions). It is also interesting to note that HPSUs funded by a public–private package of equity finance assistance also played a key role in taking new products into the global market place. The evaluation of ‘soft’ support interventions in Ireland has also been generally lacking. While individual organisations undertake internal evaluation of the impact of their business supports most are completed from a customer satisfaction
216
SMEs in a globalised world
perspective, with little evidence of the potential end outcome or impact of the various interventions on firm performance. There is no one single evaluation mechanism or set of criteria applied to ascertain the return on investment or value for money obtained from government expenditure allocated to ‘soft’ supports in terms of firm growth. In an attempt to address this gap, the Management Development Council (Forfás) are finalising a comprehensive survey of both the demand (SMEs) and supply perspective (public and private providers) of management training and development interventions to assess which interventions are most effective in developing the competencies of the owner/manager and thus facilitating SME growth. The findings of this research are scheduled for publication in 2010. Considerable resources have been devoted to supporting Irish Industrial policy and much of this has been directed towards SME policy interventions. It is however, worth acknowledging that in recent years, significant improvements have been made in terms of improving the evaluation culture in Ireland. Indeed, a report undertaken for the European Commission in 2004 (Mosselman and Prince, 2004) outlines in detail the methods employed to review the effectiveness of state aid to SMEs. In summary, the review found that evaluation practice across EU member states demonstrated a variety of approaches with many states falling below the threshold of accepted ‘best practice’ standards. In terms of Ireland, it was acknowledged that by 2004 the Irish government had a system in place whereby all state interventions were subject to evaluation. However, the study provided no insights or assessment as to the quality of these evaluations.
CONCLUSIONS AND THE WAY FORWARD The benefits of a sustained and growing SME sector to the Irish economy are manifold. SMEs are a significant component in the stock of established firms in Ireland (97 per cent of firms) and have in the main positively contributed to private sector employment (54 per cent). More specifically, this chapter has provided evidence as to the dynamics of the SME sector in Ireland. In the context of the time span of the current study, the policy environment for SMEs in Ireland, in terms of the range and diversity of supports on offer has been very active. A clear focus on SME sector emerged particularly during the 1990s, culminating in the publication of the ‘Task Force on Small Business’ in 1994 (Government of Ireland, 1994b). This represented a significant milestone in SME policy in Ireland, representing
Dynamics of the SME sector in Ireland
217
as it did the first formal recognition of the potential contribution over a range of factors of the SME sector to the Irish economy. What can we conclude about the dynamics of the SME sector in Ireland as a driver of growth in the Irish economy since 1994? A few key overriding conclusions emerge: in profiling SME activity since 1994, we see that the number of SMEs, and particularly those in the construction and service sectors have significantly increased. Recent evidence for the manufacturing sector over the period 1972–2006 has shown that small firms (less than 50 employees) and particularly micro enterprises (less than 10 employees) have significantly higher average rates of job creation than medium and larger firms and establishments (Lawless and Murphy, 2008). Further, regression analysis of the relationship between employment growth and firm characteristics in the manufacturing sector reveals that there was a negative relationship between firm size and job growth. That is, small firms create and destroy jobs at a higher percentage rate compared to larger firms (of course the absolute numbers of job gains and losses still are higher in large firms). This is not, of course, unique to the Irish economy and is consistent with international evidence (see Davis et al., 1996). However, further analysis of this dataset is required to understand the relationship between job creation rates of small firms by subperiod in the 34 years since 1972. It is impossible to comment with any certainty as to the propensity of Irish SMEs to export given the lack of available data. However, the limited evidence does indeed suggest that SMEs in Ireland are less likely to export when compared to larger firms. Based on the profiling of SMEs provided in this chapter, one would indeed tend towards a conclusion that SMEs were an important growth dynamic to the Irish economy since 1994. Most recent evidence on VAT registrations however, tends to point towards the quite surprising result that, economic growth can be experienced at a national level without any accompanying (or causing) increase in business creation activity. It is only when more data becomes available that we will be able to make a definitive assessment and suggest suitable policy interventions. The collection and maintenance of reliable databases in Ireland is imperative if serious SME development research and subsequent policy improvements are to be made. Coupled with this evaluation studies pertaining to SME interventions in the Irish economy are also lacking. It is thus extremely difficult to ascertain the effectiveness of SME policy interventions in terms of improving SME growth performance and overall impact on wider growth issues. It is worrying to find that a plethora of SME policy interventions have taken place and continue to be funded in the absence of robust evaluations of these interventions so as to ensure accountability and value for
218
SMEs in a globalised world
money and also as a means of ensuring future policy improvements. For example, a dearth of empirical-based research exists with regards to the impact of soft supports, an area of increasing expenditure by Irish development agencies. Of late, however, encouraging signs are emerging such as the fact that Forfás is undertaking a comprehensive review (the results of which are expected to be published in February 2010) of both SMEs and public and private providers of management training and development so as to attain some assessment with respect to the benefits of training to firms vis-à-vis its impact on firm growth. Thus in summary, we suggest: the need for greater evaluation metrics to assess the impact of current interventions of both a financial and a non-financial nature. In the absence of reliable and quantifiable longitudinal datasets the task is indeed a very challenging one. A Look Towards the Future of SME Development and Policies in Ireland In looking to the future, we emphasise the following key points: SMEs have the potential to play an increasingly important role in terms of contributing to the growth dynamics of the Irish economy. SMEs (compared to FDI) tend to be key contributors to local economic development and tend to be very much embedded in such localities. Of late, industrial development challenges are beginning to emerge in the context of the Irish economy, particularly with respect to the economy’s dependence on FDI. Industrial development challenges are likely to emerge in view of issues such as inter alia: has Ireland generated too much dependency on FDI from a particular source (USA), and concentrated in particular sectors? The selective and targeted industrial policy approach (to date seen as largely successful) adopted by the Irish government may face its own challenges which may be beginning to emerge. If this is the case, SMEs will be more important than ever in the future in terms of contributing to the future growth dynamics of the Irish economy in the face of issues such as increased globalisation. Recent forecasts tentatively suggest that GNP average annual growth in Ireland will be 3.8 per cent over the period 2010–2015 and 3.5 per cent over the period 2015–2020 (Fitzgerald et al., 2008b, p. ix). This compares to growth rates which averaged over 8 per cent between 1995 and 2000 (Duffy et al., 2001 p. xi). However, any forecasts would appear to be quite meaningless in the economic climate at the end of the first decade of the twenty-first century. Many challenges are likely to remain for SMEs operating within the Irish economic environment. Such challenges are likely to revolve around issues such as exporting. As a country with a relatively limited domestic market for both products and services, it is imperative
Dynamics of the SME sector in Ireland
219
that Irish SMEs who wish to expand continue to seek opportunities to grow through exporting to an increasingly globalised world. From a policy perspective, it is also necessary both to encourage and facilitate the growth of SMEs capable of achieving international growth and those firms who may not be ready for international growth but who need to strengthen their growth achieving ability in Irish markets first. In particular, SMEs operating in high growth value added sectors and in the services sector need to be strengthened. So in summary, in a marketplace that is increasingly viewed as global, we argue that increased activity on exporting and global trading is an essential strategy for the sustainability and growth of SMEs in Ireland across all sectors. It is further contended that government policy needs to generate commitment and provide for the creation of competencies needed by owner/managers to achieve higher levels of successful export activity in the SME sector as this has not been adequately targeted by SME policies to date. The evaluation of SME policy interventions is key because in learning, we are then better informed on how to assist a changing profile of SMEs (service firms, HPSUs and those providing added value products/services) who will operate in different and in an increased number of markets (international global business environment) in the future. Our final conclusion is that SMEs have indeed played an important role in transforming the Irish economy, but with a supportive and well informed policy environment (brought about as a result of good practice evaluation methodologies) they can play an even more fruitful role in the future in the face of many new challenges including that of globalisation.
NOTES 1. Defined as incomplete markets in the sense that some firms (and the evidence suggests that this tends to be smaller and younger firms) can be denied access to credit by normal lending providers (for example, banks) even though they possess a viable business project. 2. More specifically, Lenihan (1999; 2004) distinguishes between what she terms ‘partial’ versus ‘pure’ degrees of deadweight and defines deadweight as the degree to which firm projects would have gone ahead without assistance from Shannon Development in 1995. The definition incorporates the notion of degrees of deadweight as measured by time, location and scale. Lenihan et al. (2005) also argue that evaluation should take account of the other key (positive) components of additionality (that is, multipliers and linkages). 3. For more information see Wren (2007). 4. Of course, one would have to urge caution when comparing derived estimate of deadweight (or displacement for that matter) across studies due to the fact that different studies employ different methodological approaches.
220
SMEs in a globalised world
REFERENCES Acs, Z.J. and D.J. Storey (2004), ‘Introduction: entrepreneurship and economic development’, Regional Studies, 38 (8), 871–77. Acs, Z., C. O’Gorman, L. Szerb and S. Terjesen (2007), ‘Could the Irish miracle be repeated in Hungary?’, Small Business Economics, 28, 123–42. Andreosso-O’Callaghan, B. and H. Lenihan (2006) ‘Is Ireland a role model for SME development in the new member states?’, Journal of European Integration, 28 (3), 277–303. Anyadike-Danes, M., M. Hart and H. Lenihan (2010), ‘New business formation in a rapidly growing economy: the Irish experience’, Small Business Economics, 34. Audretsch, D.B. and A. R. Thurik (2001), ‘What’s new about the new economy? Sources of growth in the managed and entrepreneurial economies’, Industrial and Corporate Change, 10 (1), 267–315. Bailey, D., A. De Ruyter and N. Kavanagh (2007) ‘Lisbon, Sapir and industrial policy: evaluating the Irish experience’, International Review of Applied Economics, 21 (3), 453–67. Barkham, R., G. Gudgin, M. Hart and E. Hanvey (1996), The Determinants of Small Firm Growth: An Inter-Regional Study in The United Kingdom 1986–1990, Regional Policy and Development Series, London: Jessica Kingsley Publishers. Barry, F. (1999), Understanding Ireland’s Economic Growth, London and New York: Macmillan Press and St Martin’s Press. Burnham, J.B. (2003), ‘Why Ireland boomed’, The Independent Review, 7 (4), 537–556. Capecchi, V. (1990), ‘A history of flexible specialisation and industrial districts in Emilia-Romagna’, in F. Pyke, G. Becattini, and W. Sengenberger (eds), Industrial Districts and Inter-Firm Cooperation in Italy, Geneva: International Institute for Labour Studies, pp. 20–36. Central Bank of Ireland (1992), Annual Report, Dublin: Central Bank of Ireland. Central Bank of Ireland (1999) Annual Report, Dublin: Central Bank of Ireland. Central Statistics Office (CSO) (2006a), Annual Services Inquiry 2006, Dublin: The Stationery Office, CSO. CSO (2006b), Census of Industrial Production, Dublin: The Stationery Office. Culliton, J. (1992), ‘A time for change: industrial policy for the 1990s’, Report of the Industrial Policy Review Group, Dublin: The Stationery Office. Davis S.J., J.C. Haltiwanger and S. Schuh (1996), Job Creation and Destruction, Cambridge, MA and London: MIT Press. Drudy, P.J., (1995), ‘From protectionism to enterprise: a review of Irish industrial policy’, in A. Burke (ed.), Enterprise and the Irish Economy, Dublin: Oak Tree Press, pp. 71–89. Duffy, D., J. Fitzgerald, J. Hore, I. Kearney and C. MacCoille (2001), ‘Mediumterm review 2001–2007’, report for the Economic Social Research Institute (ESRI) Macroeconomic Forecasting Series no. 8, Dublin: ESRI. Enterprise Strategy Group (2004), Ahead of the Curve – Ireland’s Place in the Global Economy, Dublin: Forfás. Ferreira, M.L. and P. Vanhoudt (2004), ‘Catching the Celtic Tiger by its tail’, European Journal of Education, 39 (2), 209–35. Fitzgerald, J., D. Duffy and B. Bergin (2008a), ‘Medium term review 2005–2012’,
Dynamics of the SME sector in Ireland
221
report for the Economic Social Research Institute (ESRI) Macroeconomic Forecasting Series summary, Dublin: ESRI. Fitzgerald, J., A. Bergin, T. Conefrey, S. Diffeny, D. Duffy, I. Kearney, S. Lyons, L. Malaguzzi Valeri, K. Mayor and R.S.J. Tol (2008b), ‘Medium Term Review 2008–2015’, report for the Economic Social Research Institute (ESRI) Forecasting series no. 11, Dublin: ESRI. Fitzgerald, A. and M. Breathnach (1994), Technological Innovation in Irish Manufacturing Industry: Report for the Policy and Advisory Board for Industrial Development in Ireland, Dublin: Forfás. Fitzpatrick Associates Economic Consultants, (2004), Review of the Role of City and County Enterprise Boards in the Development of Micro-enterprises, Dublin: Fitzpatrick Associates. Forfás (1996), Shaping our Future – A Strategy for Enterprise in Ireland in the 21st Century, Dublin: Forfás. Forfás (2006), The Changing Nature of Manufacturing and Services – Irish Trends and International Context, Dublin: Forfás. Fritsch, M. and P. Mueller (2004), ‘Effects of new business formation on regional development over time’, Regional Studies, 38 (8), 961–75. Goodbody Economic Consultants (2002), Entrepreneurship in Ireland, Dublin: Goodbody Economic Consultants. Government of Ireland (1984), White Paper on Industry Development, Dublin: The Stationery Office. Government of Ireland (1989), Ireland: National Development Plan 1989–1993, Dublin: The Stationery Office. Government of Ireland (1994a), Ireland: National Development Plan 1994–1999, Dublin: The Stationery Office. Government of Ireland (1994b), Task Force on Small Business, Dublin: The Stationery Office. Government of Ireland (1994c), Report of the Task Force on Jobs in Services, Dublin: The Stationery Office. Government of Ireland (1994d), Operational Programme for Industrial Development 1994–1999, Dublin: The Stationery Office. Government of Ireland (1995), Small Business Operational Programme, Dublin: The Stationery Office. Government of Ireland (2000), Ireland: National Development Plan 2000–2006, Dublin: The Stationery Office. Gray, A.W. (ed.) (1997), International Perspectives on the Irish Economy, Dublin: Indecon. Gudgin, G., R. Scott, E. Hanvey and M. Hart, (1995), ‘The role of small firms in employment growth in Ireland north and south’, in J. Bradley (ed.), The Two Economies of Ireland: Public Policy Growth and Employment, Dublin: Oak Tree Press. Harding, R. (2000), ‘Venture capital and regional development: towards a venture capital system’, Venture Capital: An International Journal of Entrepreneurial Finance’, 2 (4), 289–311. Hart, M. and H. Lenihan (2006), ‘Estimating additionality and leverage: the interplay between public and private sector equity finance in Ireland (2000–2002)’, Venture Capital: An International Journal of Entrepreneurial Finance, 8 (4), 331–51. Harris W.C. (2005), ‘Secrets of the Celtic tiger: act two’, Issues in Science and Technology, 21 (4), 23–7.
222
SMEs in a globalised world
Industrial Development Authority (IDA) (1972), Regional Industrial Plans 1973– 1977, Dublin: IDA. Industrial Evaluation Unit (IEU) (1999), Evaluation of Micro Enterprise Supports Across National and Local Development Agencies, Dublin: IEU. Industrial Evaluation Unit (2000), Evaluation of the Seed and Venture Capital Scheme, Dublin: IEU. Kinsella, R., W. Clarke, D. Storey, D. Mulvenna and D. Coyne (1994), Fastgrowth, Small Firms: An Irish Perspective, Dublin: Irish Management Institute. Lawless, M. (2007), Export Activities of Irish-Owned Firms, Quarterly Bulletin 1, January, Dublin: Central Bank and Financial Services Authority Ireland. Lawless, M. and A.P. Murphy (2008), ‘Job turnover in Irish manufacturing 1972– 2006’, The Economic and Social Review, 39 (3), 235–56. Lenihan, H. (1999), ‘An evaluation of a regional development agency’s grants in terms of deadweight and displacement’, Environment and Planning C: Government and Policy, 17 (3), 303–18. Lenihan, H. (2004), ‘Evaluating Irish industrial policy in terms of deadweight and displacement: a quantitative methodological approach’, Applied Economics, 36, (3), 229–52. Lenihan, H. and M. Hart (2004), ‘The use of counterfactual scenarios as a means to assess policy deadweight: an Irish case study’, Environment and Planning C: Government and Policy, 22 (6), 817–39. Lenihan, H. and M. Hart (2006), ‘Evaluating the additionality of public sector assistance to Irish firms: a question of ownership?’, Policy Studies, 27 (2) 115–33. Lenihan, H., M. Hart and S. Roper (2005), ‘Developing an evaluative framework for industrial policy in Ireland: fulfilling the audit trail or an aid to policy development?, in Quarterly Economic Commentary, Summer, Dublin: Economic and Social Research Institute, pp. 69–85. Mason, C. and R. Harrison (1992), ‘The supply of equity finance in the UK: a strategy for closing the equity gap’, Entrepreneurship and Regional Development, 4 (4), 357–80. Mason, C.M. and R.T. Harrison (1997), ‘Business angel networks and the development of the informal venture capital market in the UK: is there still a role for the public sector’, Small Business Economics, 9 (2), 111–23. Mosselman, M. and Y.M. Prince (2004), Review of Methods to Measure the Effectiveness of State Aid to SMEs, final report to the European Commission, The Netherlands: EIM. Office of the Attorney General (1993), ‘Irish Statute Book: Industrial Development Act of 1993’, accessed 25 March 2009 at www.irishstatutebook.ie/1993/en/act/ pub/0019/index.html. Revenue Commissioners (2004), Revenue Commissioners’ Annual Report 2003, Dublin: Revenue Commissions. Ruane, F. (2004), ‘Creating a culture of evaluation – getting from rhetoric to Possibility’, paper presented to the Irish Evaluation Network, Inaugural Annual Conference 2004, A Culture of Evaluation – Exploring the Irish Experience of Evaluation, 24 September, Dublin City University. Ruane, F. and H. Görg (1999), ‘Irish FDI policy and investment from the EU’, in R. Barrell and N. Pain (eds), Investment, Innovation and the Diffusion of Technology in Europe, Cambridge: Cambridge University Press, pp. 44–67. Sapir, A., P. Aghion, G. Bertola, M. Hellwig, J. Pisani-Ferry, D. Rosati, J. Vinals
Dynamics of the SME sector in Ireland
223
and H. Wallace (2003), An Agenda for a Growing Europe: Making the EU Economic System Deliver, report for the President of the European Commission, Brussels: European Commission. Small Business Forum (2006), ‘Small business is big business’, report of the Small Business Forum, Dublin. Storey, D. (2000), ‘Six steps to heaven: evaluating the impact of public policies to support small businesses in developed economies’, in D.L. Sexton and H. Landström (eds), The Blackwell Handbook of Entrepreneurship, Oxford: Blackwell, pp. 176–94. Storey, D.J. (1994), Understanding the Small Business Sector, London and New York: Routledge. Telesis Consultancy Group (1982), ‘A review of industrial policy’, NESC report no. 64, Dublin. Thurik, R. and S. Wennekers (2004), ‘Entrepreneurship, small business and economic growth’, Journal of Small Business and Enterprise Development, 11 (1), 140–49. van Stel, A.J. and D.J. Storey (2004), ‘The link between firm births and job creation: is there a upas tree effect?’, Regional Studies, 38 (8), 893–909. Wennekers, S. and R. Thurik (1999), ‘Linking entrepreneurship and economic growth’, Small Business Economics, 13 (1), 27–55. Whitaker, T.K. (1958), Economic Development, Dublin: The Stationery Office. The World Bank Group (1999) ‘The World Bank annual report 1999’, accessed on 25 March 2009 at www.worldbank.org/html/extpb/annrep99/group.htm. Wren, C. (2007), ‘Reconciling practice with theory in the micro-evaluation of regional policy’, International Review of Applied Economics, 21 (3), 321–37.
Index access to finance 3 business start-ups Ireland 211, 214‒15 Portugal 97‒102 Ireland 211‒12, 214‒15 Acs, Z.J. 2‒3, 19, 20, 166, 203, 206 Adler, U. 68, 164 Africa 123 Aguirre, M.S. 51 Aitken, B. 165 Alberti, F.G. 182 Albino, V. 183, 186, 190, 192, 194 alliances, strategic see leader firm‒supplier relationship Almeida, P. 164 Amann, B. 87 America/Oceania 7 Andersson, S. 7 Andreosso-O’Callaghan, B. 202, 209 Antonelli, C. 6 Anyadike-Danes, M. 206‒7 Aranguren, M.J. 48 Arauzo, J.M. 24 Arellano, M. 170, 171 Arndt, S.W. 113 Arora, A. 143, 154 Asheim, B.T. 49, 51 Asia 7, 8, 203 China see separate entry associations Ireland, industry and trade 155 legal status and firm growth 36 Portugal, business 99, 100‒101, 106, 107, 110 Audretsch, D. 20, 22, 23, 24, 25, 31, 33, 206 Badía, E. 69 Bailey, D. 124, 202 Balearic Islands 37 Baltagi, B. 144
Barbosa, N. 162, 167 Barkham, R. 204 Barrios, S. 161, 163, 164, 165 Barry, F. 137, 167, 202, 203 Basque Country 37 networks see knowledge networks and small firms Becattini, G. 49, 115, 182, 187 Belgium 155 Bellandi, M. 115, 125‒6, 127 Bender, G. 192 Benetton 193, 197 Biggeri, M. 116 Blomström, M. 137, 138, 143, 153, 154 Boschma, R.A. 185, 186, 191, 196, 198 brands 126 fashion industry 67, 68, 69, 80, 84 China 123, 124, 126 flagship model 186 specialized Chinese towns 122, 123, 124 supplier‒retailer collaboration 188, 192, 193 Brazil 3 Breitzman, A. 3 BRIC economies 2 bridge leader firms 191‒2 Brusco, S. 49, 182 Buckley, R. 136, 139, 154, 155 Bulgaria 71 bureaucratic networks 186‒7, 198 Burnham, J.B. 206 ‘business angels’ 212 business start-ups family-based firms in Portugal 92‒3 bank funding 102 business associations 99, 100‒101 EU cohesion funds 88, 99, 102 events organisation industry 95‒8, 100‒101, 102 financing 97‒102
225
226
SMEs in a globalised world
furniture industry 93‒5, 96‒102 similarities and differences across two industries 96‒7 Ireland 206‒7, 211, 214‒15 foreign software companies 154‒5 Cabral, L. 162 Calof, J.L. 40 Caloffi, A. 123, 124 Calvo, J.L. 19, 22, 33, 40, 68, 69 Camuffo, A. 184 Canary Islands 22, 37 Cantwell, J. 150, 156, 165 Capecchi, V. 209 Capello, R. 194 Carbonara, N. 186, 188, 189, 191, 192, 193, 194 Carre, M.A. 19 Carton Group 191‒2, 193 Caves, R.E. 21, 137 Chaminade, C. 50 Chelan Li, L. 119 Chen, J.R. 33 Cheng, J. 115 China 3 specialized towns see specialized Chinese towns: Italian perspective spectacle frames 184 textiles and clothing sector 67, 71, 116, 121, 123‒5 tiles 183 Chrisman, J.J. 89 Christerson, B. 116 clients, periodic visits to small firms in Basque Country 57, 60 clothing and textiles China 67, 71, 116, 121, 123‒5 Italy 181, 182‒3 Spain see fashion manufacturing industry, Spanish clustering fashion industry in UK 67 industrial districts see leader firm‒supplier relationship MNEs in Portugal 162 software industry in Ireland 154 see also specialized Chinese towns Coe, N. 143
Cohen, W.M. 148 competitiveness, definition of firm 50 Cooke, P. 49, 51 cooperation and networking family-based firms in Portugal 107, 110 business associations 99, 100‒101, 106, 107, 110 fashion manufacturing industry, Spanish 80, 83, 84 Irish software industry industry and trade associations 155 Irish supermarket 188 MNEs and local enterprises 178 small firms in Basque Country 47‒8, 49, 52, 58, 62, 63‒4 cooperatives 36 coordination integration 185‒7, 196 Coro, G. 196 Correa, A. 22, 37 counterfeit products 126 Coyles 183 Crone, M. 139, 143, 149, 154 Culebras de Mesa, L.A. 69 Davenport, S. 5 Davis, S.J. 217 De Backer, K.D. 154‒5, 165 definition of globalisation 4 Dei Ottati, G. 182, 185 Delmar, F. 36 determinants of growth 19‒20, 41‒3 exports 33, 37, 40, 206 Gibrat’s Law 20, 21‒2, 42 conclusions 41‒3 data description 27‒30 empirical estimation 31‒3 empirical results 33‒7 estimation procedures 25‒7 industry characteristics, effects of 37‒41 MES (minimum efficient scale) 21, 23, 24, 25, 40, 41 group affiliation 33‒6, 40, 42 imports 33, 40 legal status 36, 40, 42 literature 20‒21 empirical evidence of Gibrat’s Law 21‒2
Index learning, theory of 21 service industries: gap in 22‒5 location 36‒7, 40‒41, 42 developing economies 137‒8 diversification 2, 5 Donnelly, R.G. 90, 91 Driffield, N. 162 Drudy, P.J. 208 Duffy, D. 218 Dwivedi, M. 199 Eastern Europe 203 economies of scale 5, 23, 24, 174, 184 buying power 191 economies of speed 3 Eng, I. 119 Enright, M.J. 113, 115, 122 entrepreneurship and inward FDI 154‒5, 161‒3 concluding remarks 177‒8 discussion of results 171‒7 modeling data 167‒9 empirical specification 165‒7 parametric approach 169‒70 results 171‒7 semi-parametric approach 170‒71 theoretical framework 163‒5 Ericson, R. 21 European Union 2, 3, 5‒8, 115, 202, 213, 216 cohesion funds 88, 99, 102 fashion industry 70‒71, 74, 83 events organisation industry see family-based firms in Portugal exchange rates 125 exports 2, 6, 7, 123‒4 export-oriented MNEs 164 Ireland 153, 156 Portugal 162, 164 firm growth and 33, 37, 40, 206 Ireland 217, 218‒19 Irish software industry 139, 143, 146 export-oriented MNE sector 153, 156 measuring absorptive capacity 148‒53 eyewear 184
227
fake labels 126 family-based firms in Portugal 87‒9 advantages and limitations of 103‒5 business associations 99, 100‒101, 106, 107, 110 business-start ups 92‒3 bank funding 102 business associations 99, 100‒101 EU cohesion funds 88, 99, 102 events organisation industry 95‒8, 100‒101, 102 financing 97‒102 furniture industry 93‒5, 96‒102 similarities and differences across two industries 96‒7 conclusions 108‒10 evolution of family-based relationships 103‒5 nepotism 91 succession 91‒2, 106‒8 theoretical perspectives 89‒92 Fariñas, J. 19, 22, 40 fashion manufacturing industry, Spanish 67‒70 challenges to EU fashion industry 70‒71 evolution of 71‒4 innovation 75 activities 75‒7 characteristics of innovative SMEs 77‒9 patterns of 80‒82 results of 79 policy implications for 83‒4 Fernandez, Z. 7 Fernández-Otheo, C.M. 47 Ferreira, M.L. 206 finance 3 business start-ups Ireland 211, 214‒15 Portugal 97‒102 Ireland 211‒12, 214‒15 fish processing see leader firm‒supplier relationship Fitzgerald, A. 211 Fitzgerald, J. 218 flagship model 185‒6, 198 Flores, R. 165 foreign direct investment (FDI), inward 1‒2
228
SMEs in a globalised world
China 113, 114, 116, 122, 123, 126 Ireland 203, 208, 218 software industry see productivity spillovers from MNE subsidiaries Portugal see entrepreneurship and inward FDI Spain 47 foreign direct investment (FDI), outward 6, 7, 8 Fotopoulos, G. 166 France 7, 70 Fritsch, M. 203 Fujita, M. 47 furniture industry Ireland 183 Italy 183‒4, 188 Portugal see family-based firms in Portugal Gallo, M.A. 90 Garibaldo, F. 194 Gatrell, J. 91 general partnerships 36 geographic location and firm growth 36‒7, 40‒41, 42 Gereffi, G. 113 Germany 7, 67‒8, 70 Gibrat’s Law 33 Geroski, P. 21, 166, 175 Giarratana, M. 135, 155 Gibrat’s Law see determinants of growth Girma, S. 146 globalisation, definition of 4 Goddard, J. 33 Görg, H. 20, 135, 137, 138, 153, 154, 164, 165 government intervention Ireland 157, 183, 207‒12, 218, 219 effectiveness 212‒16, 217‒18 software industry 155 Portugal 161, 177‒8 Grandori, A. 185, 186, 189, 192 Granovetter, M. 96 Gray, A.W. 203 group affiliation and firm growth 33‒6, 40, 42 growth of firms see determinants of growth
Gudgin, G. 204 Guelpa, F. 125 Guercini, S. 67 Haddad, M. 148 Hale, G. 138 Harding, R. 211 Härdle, W. 170 Harris, R. 137, 153 Harris, W.C. 202, 203 Harrison, B. 185, 188, 193 Hart, M. 211, 214, 215 Hart, P. 21 Havnes, E. 52 Heanue, K. 183 Hilmersson, M. 7 Hines, T. 68‒9 Hitt, W.D. 51 Hong Kong 123 hospitality sector 22 Howorth, C. 91, 105 Humphrey, J. 114, 121 Ijiri, Y. 21 imports 5 firm growth and 33, 40 incomplete markets 210‒11 India 3, 165 industrial districts see leader firm‒supplier relationship information systems small firms in Basque Country 56‒7, 60 innovation behaviour/capacity 2‒3 family-based firms in Portugal 107, 110 fashion manufacturing industry, Spanish 67‒70 challenges to EU fashion industry 70‒71 evolution of 71‒4 innovation strategy 75‒82 policy implications for 83‒4 Ireland 209, 210 software industry, measuring absorptive capacity 148‒53 leader firm‒supplier relationship 184, 186, 189, 192, 194, 197, 198 small firms in Basque Country 47‒8, 49, 50‒51, 54, 57, 61‒4
Index specialized towns in China 121‒2, 123, 124, 127 survival of firms 22 Ireland 7, 202‒4, 216‒19 business start-ups 206‒7, 211, 214‒15 construction sector 206, 217 fish processing see leader firm‒supplier relationship furniture 183 industrial policy 207‒12 effectiveness 212‒16, 217‒18 micro firms 206, 217 profile of SME activity 1994‒2005 204‒6 software industry 138‒40 MNEs see productivity spillovers from MNE subsidiaries taxation 139, 155 islands and growth of firms 22, 37 Italy 7, 67, 70, 181 evolving industrial district 182‒5, 192 furniture 183‒4, 188 spectacle frames 184 see also specialized Chinese towns, Italian perspective Jacobson, D. 183, 185, 196 Javorcik, B. 137, 138 Johanson, J. 6 joint stock companies 36, 40, 42 Jones, R.M. 67 Jovanovic, B. 21, 31, 33 Kathuria, V. 165 Kelly, K. 188 Kinsella, R. 204 Kirchhoff, B.A. 3 knowledge intensive export-oriented MNE sector 153, 156‒7 knowledge management small firms in Basque Country 47‒8, 51, 54, 56‒7, 60, 62 knowledge networks and small firms 47‒8 competitiveness: key issues 49‒50 characteristics of small firms 49‒50 cooperation networks 49, 52
229
firm competitiveness 50 innovation capacity 49, 50‒51 knowledge management 51 management models 51‒2 conclusions 63‒4 empirical analysis 52‒3 research methodology 53‒6 empirical results 56‒63 cooperation attitude and networking 58 definition of profiles for policymakers 58‒60 information systems and knowledge management 56‒7 information systems quality 60‒61 innovation process 57, 61‒3 management tools used 57 size specific features 56 see also leader firm‒supplier relationship Kokko, A. 138, 142, 144, 148, 153 Koschatsky, K. 194 Krugman, P. 47 La Croix, S.J. 3 Langlois, R.N. 196, 198 Lawless, M. 206, 207, 217 Lazarova, S. 26 Lazerson, M.H. 184, 193 Le Pechoux, B. 69 leader firm‒supplier relationship 181‒2 case study background 187‒9 codesign practices 192‒3 degree of dependency 193‒4 frequency of communication 190 information sharing 191‒2 partnership 189‒90 relationship impact 194‒6 resource sharing 190‒91 conclusion 197‒8 contextual environment 196‒7 evolving industrial districts 182‒5 mechanisms 185‒7 legal status and firm growth 36, 40, 42 Lenihan, H. 210, 212, 213‒14, 215 Levitt, T. 5 limited liability companies 36, 40 limited partnerships 36 Lin, P. 164
230
SMEs in a globalised world
Lipsey, R.E. 137 Liu, Z. 154 location and firm growth 36‒7, 40‒41, 42 Long, Z. 116, 121 Lorenzen, M. 188 Lotti, F. 40 Lundvall, B-A. 50 Luxottica 184 Ma, L.J.C. 117 McGrath, H. 185, 186, 188, 189, 192, 193, 196, 197 Machado, J.A.F. 40 McKeon, H. 165 Mansfield, E. 21 manufacturing sector 47, 217 entrepreneurship and inward FDI 171‒7 fashion see fashion manufacturing industry, Spanish firm growth 20, 22‒7, 37‒41, 42 data description 27‒30 empirical estimation 31‒3 empirical results 33‒7 location 36‒7, 42 MES (minimum efficient scale) 23, 24, 25, 40, 41 scale economies 23, 24 Ireland 205‒6, 207, 209, 217 productivity spillovers from MNE subsidiaries 137, 154‒5, 165 market failure 210‒11 market niches 3, 5, 182 Markusen, J. 161, 163 Maskell, P. 194 Mason, C. 211 Mata, J. 162, 166 Microsoft 138 minimum efficient scale 21, 23, 24, 25, 40, 41 Montes, V. 68 Mosselman, M. 216 Mowery, D.C. 148 multinational enterprises (MNEs) China 113, 114, 116, 122, 123, 126 growth of firms and group affiliation 33‒6, 40, 42 local entrepreneurial activity and FDI 154‒5, 161‒3, 177‒8
modeling 165‒71 results 171‒7 theoretical framework 163‒5 production systems 114 productivity spillovers from MNE subsidiaries 135‒8 conclusions 156‒7 data and methodology 140‒44 findings 144‒53 reasons for 153‒5 software industry in Ireland 138‒40 transfer pricing 139 value chains 114 Mundim, A.P.F. 5 Musgraves 193 Myro, R. 47 Mytelka, L.K. 2 Narula, R. 5 Natuzzi 183‒4 Nayaran, D. 52 Netherlands 7, 22, 67 networks small firms and knowledge 47‒8 conclusions 63‒4 empirical results 56‒63 key issues for competitiveness 49‒53 research methodology 53‒6 see also cooperation and networking niche markets 3, 5, 182 Nonaka, I. 51 Nooteboom, B. 19, 51 Oceanpath see leader firm‒supplier relationship O’Gorman, C. 141 oligopoly/oligopsony 185, 193, 197 Oliveira, B. 22‒3, 33 Oracle 138 O’Riain, S. 143 outsourcing 7 family-based firms 105 fashion industry 67, 68 Pakes, A. 21 Paniccia, I. 182 patents 3
Index Pavitt, K. 161, 174 Pfaffermayr, M. 33 Phelps, E. 97 Piore, M. 182 Plummer, L.A. 203 Poland 70, 71 Porter, M.E. 5, 49 Portugal 7 family-based firms see separate entry fashion industry 70 Gibrat’s Law 22‒3, 33 inward FDI see entrepreneurship and inward FDI Prais, S.J. 21 preference shares 211 productivity spillovers from MNE subsidiaries 135‒8, 164‒5 conclusions 156‒7 data and methodology 140‒44 findings absorptive capacity 148‒53 productivity growth 146‒8 productivity levels 144‒6 reasons for 153‒5 software industry in Ireland 138‒40 transfer pricing 139 protectionism 2, 208 public goods 127 Qiu, H. 116 relational contracting 186‒7, 198 research and development expenditure 3, 5, 210 Irish software industry 143, 146, 148, 154 measuring absorptive capacity 148‒53 leader firm‒supplier relationship 184‒5 Resende, M. 33 Rivera-Batiz, F.L. 161 Robbins, D.K. 19 Robinson, D.T. 6 Robinson, P. 170 Romania 71 Ruane, F. 203, 212 Rugman, A.M. 185, 186 Russia 123 Russo, M. 183
231
Sabel, C.F. 183 Salas Fumas, V. 50 Sammarra, A. 182 Sapir, A. 8, 202 scale economies 5, 23, 24, 174, 184 buying power 191 Scheffer, M. 67 Schmitz, H. 124, 187 Schmögnerová, B. 2 Scott-Kennel, J. 141 seed capital 211, 215 Segarra, A. 24, 25, 29 Sengenberger, W. 187 services sector 3 entrepreneurship and inward FDI 171, 172, 173‒4, 177 firm growth 20, 22‒5, 37‒41, 42 data description 27‒30 empirical estimation 31‒3 empirical results 33‒7 location 36‒7, 42 MES (minimum efficient scale) 21, 23, 24, 25, 40, 41 scale economies 23, 24 Ireland 205, 206, 209, 217 Sharma, P. 89 Smith, K. 50 social capital 52 social networks 186‒7, 189 sofa district of Matera 183‒4 software industry see productivity spillovers from MNE subsidiaries South East Asia 123 Spain 7, 165 Basque Country 37 networks see knowledge networks and small firms fashion see fashion manufacturing industry, Spanish firm growth see determinants of growth specialization 3, 6 specialized Chinese towns: Italian perspective 113‒14 case studies 123‒5 challenges to Italian districts 125‒7 conclusion 127‒8 four types of 119‒22 local development and 114‒16
232
SMEs in a globalised world
multivariate analysis 116‒19 strategies, business 126‒7, 128 spectacle frames 184 Stengg, W. 68 Storey, D.J. 211, 212 strategic alliances see leader firm‒supplier relationship strategic plans small firms in Basque Country 56, 60, 63, 64 Superquinn see leader firm‒supplier relationship suppliers see leader firm‒supplier relationship Sutton, J. 21 Sweden 7 SWOT (strengths, weaknesses, opportunities and threats) analysis 50, 60 opportunities and threats/challenges for SMEs 4‒5 Symantec 138 tacit knowledge 51, 56‒7, 183 Taiwan 123 Taplin, I.M. 68 taxation 139, 155, 208 textiles and clothing China 67, 71, 116, 121, 123‒5 Italy 181, 182‒3 Spain see fashion manufacturing industry, Spanish Thurik, R. 203, 206 tiles, ceramic 183 total quality management 64
transaction costs 105, 185 transfer pricing 139 transition economies 137‒8 Tungate, M. 69 United Kingdom 7, 67, 70, 139, 165 Gibrat’s Law 21, 26‒7, 33 United States 138, 139, 218 Gibrat’s Law 21 van Stel, A. 2, 203 Venezuela 165 venture capital funds 211, 212, 214‒15 Verheul, I. 5 Verspagen, B. 148, 153 vertical integration 185, 196, 198 Voulgaris, F. 33 Walcott, S.M. 119 Walz, U. 155 Wang, J. 116, 119 Watts, G. 89 Wei, Y. 138 Wennekers, S. 206 Westhead, P. 90 Whitaker, T.K. 208 Williamson, O.E. 185 World Trade Organization (WTO) 6 Agreement on Textiles and Clothing (ATC) 67, 71, 74, 83 Yao, Y. 117 Yeung, G. 119 Zahra, S.A. 89, 90