Differences in agglomeration externalities and industrial regimes between locations generate performance differentials for their localized economic activities. For more than two decades, scholars have debated which externality is dominant for growth and under which regime. The present study aims to resolve this debate by analysing the influence of agglomeration economies on the growth of five-digit manufacturing sectors and firms in Indonesia between 2000 and 2009 discriminating cities and regencies. Specialization, competition, population density, human capital, and a set of varieties are employed. This is conducted shedding the light on policy implications of economic variety sectoral decomposition functional to revitalize Indonesian manufacturing growth after the Asian Financial Crisis, which substantially hits the Indonesian economy and manufacturing. Empirical evidence reveals that Indonesian policymakers should develop initiatives to support the competitiveness of key labour-intensive industries and manufacturing transformation towards knowledge-based productions. This can be achieved through promoting key specialised clusters characterized by large sectoral interconnectivity favouring inter and intra-industry knowledge spillovers, which allow underpinning the competitiveness of clusters and overcoming the two typical drawbacks of highly specialized locations (lock-in and lack of resilience). The formation of human capital, and the development of technologically advanced industries come to light as crucial drivers to construct a more conductive innovative environment and reduce manufacturing exposure to external industry-specific shocks. Population density and industrial diversity antithetically influence manufacturing growth in cities and regencies due to their economic heterogeneities.
Contents
List of Figures
List of Tables
List of Abbreviations
1 Introduction
1.1 Academic and policy contributions
1.2 Chapters’ outlines
2 New Economic Geography’s peculiarities and limitations
2.1 Introduction
2.2 The main characteristics and critics of the New Economic Geography
2.3 Agglomeration formation and development under NEG
2.4 Conclusions
3 The role of agglomeration externalities on economic growth
3.1 Introduction
3.2 Technological externalities
3.2.1 Knowledge exchange within specialized clusters and its shortcomings
3.2.2 The contributions and challenges of economic variety decomposition
3.2.2.1 Knowledge transmission across sectors
3.2.2.2 Portfolio diversification effect
3.2.2.3 Policy implications of economic variety sectoral decomposition
3.3 Competition externalities
3.4 Path-dependency mechanism of agglomeration externalities
3.5 Conclusions
4 Measuring agglomeration externalities and clustering identification
4.1 Introduction
4.2 Agglomeration externalities measures
4.2.1 Location quotient index
4.2.2 Competition index
4.2.3 Economic varieties decomposition
4.3 Detecting spatial clustering
4.3.1 Spatial autocorrelation
4.3.2 Spatial weight matrix
4.3.3 Global Moran’s I index of spatial dependence
4.3.4 Local spatial autocorrelation
4.4 Conclusions
4.A Appendix: The entropy decomposition theorem
5 Economic and policy transformations, and manufacturing revitalization challenges
5.1 Introduction
5.2 Indonesian economy and policy evolutions
5.2.1 The New Order Regime (1966-1998)
5.2.1.1 Rehabilitation and stabilisation (1967-1972)
5.2.1.2 Intervention and protectionism (1973-1981)
5.2.1.3 Rationalization and export orientation (1982-1996)
5.2.2 The Asian Financial Crisis (1997-1998)
5.2.3 Following the Asian Financial Crisis (1999-onwards)
5.2.3.1 Recovery (1999-2003)
5.2.3.2 Regional and industrial cluster policies (2004- onwards)
5.3 The rise and fall of manufacturing and the beginning of its transformation
5.3.1 Manufacturing transformation
5.4 The current manufacturing challenges
5.4.1 Innovation environment
5.4.2 Human capital formation
5.4.3 Urbanization trajectory
5.5 Conclusions
6 Data collected, manufacturing composition change and agents’ localization heterogeneity
6.1 Introduction
6.2 Data collected
6.3 Share of large and medium enterprises within manufacturing . . .
6.4 Sectoral composition change within large and medium manufacturing industries
6.5 Trajectories of technology intensity industries and five-digit sectors within large and medium manufacturing operations
6.6 Agents’ localization between cities and regencies
6.6.1 Cities and regencies heterogeneity
6.7 Conclusions
6.A Appendix: Denominations of the twenty highest and lowest five- digit sectors growth
6.B Appendix: The Independent Samples t-test
7 The influence of agglomeration externalities on established manufacturing growth
7.1 Introduction
7.2 Empirical specification
7.3 Data construction and descriptive statistics
7.4 Estimation results
7.5 Conclusions
8 The dynamic impact of agglomeration externalities on manufacturing structure
8.1 Introduction
8.2 Empirical specification
8.3 Data construction and descriptive statistics
8.4 Empirical results
8.4.1 The impact of agglomeration externalities on manufacturing structure
8.4.2 The effect of agglomeration forces on technological relatedness
8.4.3 The role of (un)related variety on two-digit sectors
8.4.4 Robustness check: Endogeneity
8.5 Conclusions
8.A Appendix: The impact of (un)related variety on labour productivity of two-digit sectors
9 Key embedded specialised clusters as drivers for growth
9.1 Introduction
9.2 Data construction and descriptive statistics
9.3 The persistent presence of hotspots
9.4 Discovering key embedded specialized clusters in Eastern Jakarta
9.5 Spatial clustering of industrial development
9.6 Conclusions
9.A Appendix: Sectors of Eastern Jakarta in 2000 and 2009
10 Conclusions, policy implications, and new research agenda
10.1 Introduction
10.2 Main findings and policy implications
10.3 Policy framework for manufacturing revitalization
10.4 Limitations and agenda for further research
List of References
Author’s Biography
Roberto earned his Bachelor and Master degree in Italy where he was awarded of various scholarships to study in Sligo (Ireland), New Paltz and New York City (USA). After his graduation in Italy and before embarking to his Ph.D. in UK, Roberto worked for several multinational enterprises based in Italy and SouthEast Asia, where he also cooperated with several international universities and colleges teaching numerous business modules for four years.
The present work has produced several conference papers and presentations within referred and ordinary sessions across Europe (including UK) as well as at the Pacific Regional Science Conference Organisation (PRSCO, biannual) in Bandung, Indonesia. Applying the current conceptual and empirical framework to Indonesia, Vietnam and Italy in order to unfold the impact of agglomeration externalities on their manufacturing expansion. During my Ph.D, I had the pleasure to be the chairman of two conference sessions, I received the possibility to attend the summer school in Poland sponsored by the European Regional Science Association (ERSA), and a conference bursary by the Regional Study Association (RSA). In addition, the present work has generated an academic article in the Growth and Change Journal.
Abstract
Differences in agglomeration externalities and industrial regimes between locations generate performance differentials for their localized economic activities. For more than two decades, scholars have debated which externality is dominant for growth and under which regime. The present study aims to resolve this debate by analysing the influence of agglomeration economies on the growth of five-digit manufacturing sectors and firms in Indonesia between 2000 and 2009 discriminating cities and regencies. Specialization, competition, population density, human capital, and a set of varieties are employed. This is conducted shedding the light on policy implications of economic variety sectoral decomposition functional to revitalize Indonesian manufacturing growth after the Asian Financial Crisis, which substantially hits the Indonesian economy and manufacturing. Empirical evidence reveals that Indonesian policymakers should develop initiatives to support the competitiveness of key labour-intensive industries and manufacturing transformation towards knowledge-based productions. This can be achieved through promoting key specialised clusters characterized by large sectoral interconnectivity favouring inter and intra-industry knowledge spillovers, which allow underpinning the competitiveness of clusters and overcoming the two typical drawbacks of highly specialized locations (lock-in and lack of resilience). The formation of human capital, and the development of technologically advanced industries come to light as crucial drivers to construct a more conductive innovative environment and reduce manufacturing exposure to external industry-specific shocks. Population density and industrial diversity antithetically influence manufacturing growth in cities and regencies due to their economic heterogeneities.
Key words: Agglomeration externalities, related and unrelated varieties, Indonesian economy and policy evolutions.
Jel classification: D62, O18, O25, R11
List of Figures
1.1 Structure and key contents of the present study
2.1 The main NEG’s peculiarities and critics
2.2 The main NEG’s agglomeration forces linked through the circular cumulative causation
3.1 The diversification process of complementary competences accumulation and recombination into new and related pathways of growth
3.2 The path-dependency mechanism of economic configuration
4.1 The Queen contiguity conceptualization
4.2 The four spatial associations of LISA statistics
5.1 The annual GDP growth in Indonesia, East Asia & Pacific (developing only), and the world’s economy by Indonesian economic phases and policy interventions during 1961 and 2013
5.2 The annual percentage growth of GDP, GDP per capita, export and import of goods and services, and FDI inflows in Indonesia between 1961 and 2013
5.3 The inflation rate variations in Indonesia between 1961 and 2013
5.4 The structural change in the Indonesian economy between 1960 and 2013: Value added and employment as a percentage of GDP and total employment respectively of agriculture, industry, service and manufacturing
5.5 The map of six Indonesian economic corridors set by the Master Plan for the Acceleration and Expansion of Indonesia’s Economic Development 2011-2025
5.6 Annual average of manufacturing value added and GDP growth, and manufacturing import and export as a percentage of merchandise by the country’ stages between 1967 and 2013
5.7 The contribution of major manufacturing industries as a percentage of total manufacturing value added in Indonesia between 1990 and 2009.
5.8 Research and development expenditure in Indonesia and East Asia and Pacific (developing only) as a percentage of GDP in 2000 and 2009
5.9 Average of patent’s applications of non-residents and residents by different stages in Indonesia between 1967 and 2013 85
5.10 High-technology exports in Indonesia as a percentage of manufactured exports between 1989 and 2012 85
5.11 Average of education expenditure and public spending in Indonesia and East Asia and Pacific (developing only) by different stages between 1988 and 2012 86
5.12 Labour force, school enrolment, and unemployment with secondary and tertiary educations in Indonesia between 2000 and 2008 87
5.13 Human Development Index (HDI) by provinces in Indonesia between 1996 and 2012 87
5.14 Population density, urban and rural population in Indonesia between 1960 and 2012 89
5.15 The average growth of population, urban and rural areas and population’s concentration within major Indonesian cities by different stages between 1967 and 2013 89
5.16 The average population growth by Indonesian provinces between 2000 and 2010 90
6.1 Employment and value added share of large and medium enterprises within manufacturing in Indonesia between 2000 and 2012 95
6.2 Employment, value added, labour productivity (log scale) of technology intensity clusters within large and medium manufacturing enterprises between 2000 and 2009 103
7.1 The average of annual employment, value added and labour productivity growth rates at the firm-level between 2000 and 2009 disaggregated by types of locations, technology intensity industries and firm’ sizes 128
9.1 Employment and value added distribution in 2000 (top-left and bottom-left respectively), the standard deviation of average annual employment and value added growth between 2000 and 2009 (top- right and bottom-right respectively) 174
9.2 Average annual employment and value added growth within cities in and out of Java 175
9.3 Average annual employment and value added growth within regencies in and out of Java 176
9.4 The persistent presence of hotspots clusters (High-High values) between 2000 and 2009 178
9.5 Local annual average of employment and value added growth within the clusters GSP and JB 180
9.6 Five-digit industrial configuration of Eastern Jakarta in 2000 186
9.7 Five-digit industrial configuration of Eastern Jakarta in 2009 187
9.8 Bivariate LISA statistics of LQ in 2000 and the spatial lag of RV in 2009 (top), and RV H MH in 2000 and the spatial lag of V ARI ET Y in 2009 (bottom) 192
List of Tables
6.1 Two-digit sector shares of establishments, employment and value added in 2000 and their variations within large and medium manufacturing 99
6.2 The twenty-highest five-digit sectors of average annual growth rates (%) in terms of number of establishment, employment, value added and labour productivity within large and medium manufacturing between 2000 and 2009 104
6.3 The twenty-lowest five-digit sectors of average annual growth rates (%) in terms of number of establishments, employment, value added and labour productivity within large and medium manufacturing between 2000 and 2009 105
6.4 Nomenclature of variables and descriptive statistics disaggregated by cities and regencies, and the Independent Samples t-test 111
6.5 The Independent Samples t-test of employment, value added, and labour productivity of two-digit sectors between cities and regencies. . 112
6.6 Nomenclature of the five-digit sectors presented in Table 6.2 and Table 6.3 114
7.1 Nomenclature of variables and their means and standard deviations disaggregated by cities and regencies 127
7.2 The influence of agglomeration externalities on average annual employment growth of five-digit sectors and firms disaggregated by cities and regencies between 2000 and 2009 132
7.3 The influence of agglomeration externalities on average annual value added growth of five-digit sectors and firms disaggregated by cities and regencies between 2000 and 2009 133
7.4 The influence of agglomeration externalities on average annual labour productivity growth of five-digit sectors and firms disaggregated by cities and regencies between 2000 and 2009 134
8.1 Nomenclature of variables and their descriptive statistics of five-digit sectors and firms disaggregated by cities and regencies 143
8.2 Agglomeration externalities impact on five-digit sectors and firms’ employment in cities and regencies between 2000 and 2009 150
8.3 Agglomeration externalities impact on five-digit sectors and firms’ value added in cities and regencies between 2000 and 2009 151
8.4 Agglomeration externalities impact on five-digit sectors and firm’s labour productivity in cities and regencies between 2000 and 2009. . . 152
8.5 Agglomeration externalities effect on five-digit sectors’ value added disaggregated by technology intensity industries within cities and regencies between 2000 and 2009 155
8.6 The influence of (un)related variety on employment by two-digit sectors within cities and regencies between 2000 and 2009 159
8.7 The influence of (un)related variety on value added by two-digit sectors within cities between 2000 and 2009 160
8.8 The influence of (un)related variety on value added by two-digit sectors within regencies between 2000 and 2009 161
8.9 The impact of (un)related variety on labour productivity by two-digit sectors within cities between 2000 and 2009 167
8.10 The impact of (un)related variety on labour productivity by two-digit sectors within regencies between 2000 and 2009 168
9.1 Nomenclature, descriptive statistics and the independent samples t- test of local aggregated variables 173
9.2 The two-digit industrial configuration of Eastern Jakarta and its evolution between 2000 and 2009 185
9.3 Two-digit average annual employment and value added growth, employment share, industrial linkages, and five-digit specialization within the clusters JB and GSP 191
9.4 Five-digit denominations reported in Figure 9.6 and Figure 9.7 195
List of Abbreviations
Abbildung in dieser Leseprobe nicht enthalten
1. Introduction
Indonesia is one of the largest and stable economies in Asia characterized by abundant natural resources such as mineral fuels, lubricants, animal and vegetable oils, fats, and waxes. It is the first South-East Asia country to become a member of the G-20 major economies since 1999, when the forum was established (see, for instance, Hermawan, Sriyuliani, Hardjowijono, & Tanaga, 2011). Indonesia is also a co-founder member of the Association of Southeast Asian Nations (ASEAN)1 established in 1967 for political and economic cooperation among Southeast Asian countries. Indonesia recently witnessed deep transformations in terms of industrial scale and structure, urban concentration and socio-economic conditions. These mutations were mainly dictated by the Asian Financial Crisis (AFC, 1997-1998) that hits Indonesia economy and manufacturing activities at much higher pace than other developing economies in the region highlighting the country’s weaknesses to external shocks. Indonesia struggled to recover since the economic level pre- crisis is not yet reached. Despite this, it became one of the most dynamic economies in the region.
During the period beginning in 2000 and ending in 2009, Indonesian GDP grew annually between 4% and 6%, GDP per capita increased between 2% and 5%, and the population expanded between 1% and 2% (World Bank, 2015). People living within urban centres accounted for 42% in 2000 and 49% in 2009 of the total population (almost a quarter of a billion), and more than one half lived in urban areas in 2011 (World Bank, 2015). These favourable economic conditions encouraged Indonesian industries to re-focus on their domestic markets. The exportation of goods and services, as a percentage of GDP, markedly decreased from 41% in 2000 to 24% in 2009. Manufacturing exports, as a percentage of merchandise exported, decreased from 57% in 2000 to 41% in 2009 (World Bank, 2015). Exports of high technology industries declined from 16% in 2000 to 13% in 2009 of total manufacturing exports (World Bank, 2015). Despite these contractions, manufacturing experienced significant growth in terms of value added and labour productivity though employment grew at much lower rates between 2000 and 2009. This can be due to manufacturing transformations (e.g. the adoption of new technologies) in order to cope macro- economic mutations and the increasing intensity of domestic and international competitions. However after the AFC, manufacturing began to growth at lower pace than the overall economy showing a decline trajectory and a potential threat of deindustrialization emerged. The substantial importance of few labour-intensity industries within manufacturing structure exposed it to external shocks, as a result, a reduction of their competitiveness undermined the overall manufacturing growth in Indonesia.
Manufacturing plays an important role for the economic growth in Indonesia due to its high productivity and propensity to cluster generating agglomeration externalities. The localization of manufacturing has not spread all over Indonesia, but it clustered in certain locations such as the cities of Jakarta (with particular reference to its Northern and Eastern areas), Tangerang, Bandung and Surabaya, and the regencies of Tangerang, Bogor, Bekasi, and Bandung, among others. Although Java Island is characterized by the highest concentration of manufacturing within the country, Indonesia witnessed a diversification of manufacturing growth between 2000 and 2009, where less dense locations grew faster than others. These differences in economic structure and growth between Indonesian locations generated performance differentials for their localized economic activities. Numerous industries show higher performance within regencies, whereas others are more productive within cities2, which are characterized by diverse economic configuration. Thus, some questions emerge: Why certain economic activities have higher growth in certain places and under a certain industrial regime? What are the determinants of such growth? A large body of literature has been made in order to explain these questions; theoretical and empirical literature point out that firms and workers have higher productivity within large and dense economic environments (see, for instance, Melo, Graham, & Noland, 2009; Puga, 2010; Rosenthal & Strange, 2004). This can be associated with the proximity effect of economic activities, from which rises agglomeration externalities.
However, there is a little agreement among researchers of which externalities, specialized (Glaeser, Kallal, Scheinkman, & Shleifer, 1992; Marshall, 1890), or diversified (Bairoch, 1988; Jacobs, 1969) play a predominant role for innovation and growth. The impact of agglomeration externalities can also differ across sectors and space due to their heterogeneity (see, for instance, De Groot, Poot, & Smit, 2009; Rosenthal & Strange, 2004; Van Oort, 2007). Besides this, scholars debate under which market structure innovation is optimized (see, for instance, Beaudry & Schiffauerova, 2009; De Groot et al., 2009). A more recent vein of literature refers to the Darwinian selection of firms as competition pushes weaker economic activities out of the market where the most efficient and innovative firms survive enhancing their performance and the relative aggregations (i.e. sectors and locations) (Combes, Duranton, Gobillon, Puga, & Roux, 2012; Duranton & Puga, 2003; Melitz, 2003). This debate has been alimented over time since scholars have found evidence to support diverse conceptualizations (see, for instance, De Groot et al., 2009; De Groot, Poot, & Smit, 2015). A potential cause of this inconclusive debate stems from the misspecification of economic variety (Boschma, Minondo, & Navarro, 2012).
This study aims to resolve this long-term academic debate testing the influence of urbanization, specialization, competition, and a set of varieties employing the economic variety sectoral decomposition as proposed by Frenken, van Oort, and Verburg (2007). This latter allows decaying general variety without any sectoral linkages into unrelated and related varieties in order to evaluate more accurately their idiosyncratic effects on growth associated with portfolio diversification (Conroy, 1974, 1975) and inter-industry knowledge spillovers (Jacobs, 1969) respectively. The Indonesian industrial classification (KBLI 2005) and the technology intensity classification (OECD, 2011) are employed to determine the cognitive proximity among sectors. These agglomeration externalities are assessed on the expansion of large and medium five-digit manufacturing sectors and firms in terms of employment, value added and labour productivity analysing separately cities and regencies between 2000 and 2009. The present study becomes particular relevant considering the current policymakers’ challenges to revitalize manufacturing in Indonesia. It will be argue that the economic variety sectoral decomposition can provide valuable insights for policy design to bring back on track manufacturing. To the best of my knowledge, no similar studies have been conducted in Indonesia and the decomposition of economic variety has been applied to developed economies (see, for instance, Bishop & Gripaios, 2010; Boschma & Iammarino, 2009; Boschma, Minondo, & Navarro, 2011; Frenken et al., 2007; Quatraro, 2010). The rest of this chapter is organized as follows. In Section, 1.1, the academic and policy contributions of the present study are briefly presented. In Section 1.2, the flow of knowledge disaggregated by chapters is illustrated.
1.1 Academic and policy contributions
The present study aims to contribute to the existing theoretical and empirical literature in several areas, which stem from employing the economic variety sectoral decomposition, agglomeration externalities tested, the level of data employed, the country of analysis, and considering the heterogeneity between cities and regencies. They are addressed in order to provide recommendations to policymakers aiming to revitalize manufacturing in Indonesia. These contributions are schematically presented as follows.
More appropriate theoretical foundations and tailor-made industrial policies.
Decomposing economic variety into (un)related variety based on sectoral linkages allows addressing the misspecification of Jacobian externalities contributing to resolve the long-term academic debate on which externality is more predominant for growth (see, for instance, Beaudry & Schiffauerova, 2009; De Groot et al., 2009; Feldman & Audretsch, 1999; Van der Panne, 2004). The idiosyncratic roles of inter-industry knowledge spillover (Jacobs, 1969) linked to related variety, and portfolio diversification (Conroy, 1974, 1975) associated with unrelated variety can be separately assessed. This allows discerning Jacobian externalities and portfolio diversification notions along with urbanization externalities, where this latter is connected to the market-size effect á la Krugman (1991a,1991c) rather than inter-industry knowledge spillover.
Identifying large and small sectoral cognitive proximity allows tailor-made industrial policies towards relatedness and diversification enhancing economic growth and resilience. Promoting key industries with large intersectoral linkages consents to reduce the risk associated with lock-in effect and lack of economic resilience, which are typical drawbacks of having a location highly specialized. Since new external knowledge can flow between interconnected economic activities with diverse but complementary know-how, which can also generate the formation of regional (un)related branches driving to new pathways of growth where the pre-existing local economic configuration can affect their genesis. In addition, identifying local degree of heterogeneous configuration provides valuable information for policy strategies to increase embedded relatedness and/or further diversification enhancing local resilience and more balanced growth. Policymakers often ignore this relationship between growth and stability for regional economic development.
Agglomeration externalities tested and their post-impact on growth. Following the seminal work of Glaeser et al. (1992), and subsequently numerous other works (see, for instance, De Vor & De Groot, 2010; Henderson, 1997, 2003), the influence of urbanization, specialization, competition and economic diversity (general varieties3 ) are assessed on manufacturing expansion within Indonesian locations. Decomposing economic variety using entropy formula as proposed by Frenken et al. (2007) permits to assess urbanization and MAR externalities along with general variety in order to compare the empirical results with previous studies’ outcomes; and extending them through the disaggregation of general variety into unrelated and related varieties. Considering agglomeration externalities, the notion of path- dependency is implicitly embraced, which is often neglected by researchers (see, for instance, De Groot et al., 2009, for a review of thirty-one studies). Thus, the post-impact of agglomeration externalities is tested on manufacturing growth since its expansion is the result of prior efforts.
Capturing micro variations on three manufacturing growth dimensions.
The micro foundation of agglomeration externalities is considered employing the lowest sectoral digit level (five-digit) within the Indonesian industrial classification (KBLI 2005) and the single economic activity, which allow assessing micro variations and avoiding sectoral aggregation bias. This is often ignored by researchers (see, for instance, De Groot et al., 2009), which might cause a potential estimation bias where the most disaggregated level generates the most consistent economic theories. The influence of agglomeration externalities is tested on three dimensions of manufacturing growth in order to determine more precisely their idiosyncratic influence on manufacturing expansion. This is particular relevant in Indonesia considering that manufacturing value added and labour productivity grew faster than the numbers of employees during 2000 and 2009. Therefore considering the only employment dimension within the empirical analysis, manufacturing growth is not properly captured in Indonesia, which is further addressed taking into account value added and labour productivity.
The country development and discriminating cities and regencies. Most scholars (see, for instance, Bishop & Gripaios, 2010; Boschma & Iammarino, 2009; Boschma et al., 2011; Castaldi, Frenken, & Los, 2014; Frenken et al., 2007; Hartog, Boschma, & Sotarauta, 2012; Quatraro, 2010) tested the reconceptualization of economic variety within developed economies. The economic variety decomposition applied to Indonesian can provide valuable insights for researchers and policymakers due to its fast expanded economy and the current policymakers’ challenges to revitalize manufacturing. Besides this, cities and regencies show heterogeneity in terms of area size, industrial composition, population density and availability of skilled workers determining the generation and magnitude differentials of agglomeration externalities, which lead to unlike performance of their localized economic activities.
Established manufacturing sectors show higher performance within regencies characterized by lower competition and cost of factors of production; whereas established firms and the overall manufacturing structure are more productive within Indonesian cities denoted by large local demand, heterogeneous industries, availability of skilled workers, and the localization of high and medium-high technology intensity industries. Discerning urban environments and wider geographical scales allows to take into account for their heterogeneity in terms of economic configuration and performance enhancing inference and policy relevance between these two diverse types of administrative units. Neglecting for their idiosyncratic differences analysing the entire country indiscriminately can lead to erroneous outcomes as shown in the work of Ercole and O’Neill (forthcoming). This has implications on previous study’ findings (see, for instance, De Groot et al., 2009) that merely analyse agglomeration externalities at the country or regional-level since they need to be interpreted carefully.
Manufacturing decline and its revitalization. The aforementioned framework is employed to manufacturing expansion in the post-shock period. The two- year shock (AFC, 1997-1998) determined a series of changes in Indonesia such as the end of the Soeharto’s authoritarian regime favouring new economic reforms, and the increasing of labour costs becoming one of the countries with the highest minimum wages in the world on average (OECD, 2012). Besides this, new trade agreements in the region increased international and domestic rivalry, with particular regard to countries (e.g. Vietnam and Cambodia) characterized by lower cost of productions in comparison of Indonesia. These mutations caused a substantial decline of manufacturing competitiveness with particular reference to labour-intensity industries, and manufacturing struggled to ”bounce back”. This led to manufacturing transformation towards more competitive and innovative activities, especially favouring knowledge-based productions.
Policymakers’ challenges emerged aiming to support manufacturing competitiveness and its transformation where innovation and human capital come to light as key drivers to lead to a second period of industrialization in the country. It will be argued that the identification of economic relatedness and heterogeneous configuration within locations can provide valuable insights for Indonesian policymakers. Since this allows developing ad-hoc policies strategies in order to prioritise specialized clusters characterized by large sectoral interconnectedness and to enhance manufacturing diversification increasing economic growth and stability. These new insights can be embedded within recent Indonesian policies, which began to support key clusters focusing on critical issues for manufacturing growth such as innovation, human capital and spatial inequalities adopting location and cluster approaches, which recognize local heterogeneity and the important role-played by agglomeration externalities for local growth. In this framework conditions, the present study becomes particular relevant in Indonesia since its economy progressively moves towards a knowledge-based economy where the learning process is playing an increasing role for employment and productivity growth (Menkhoff, Evers, Wah, & Fong, 2011).
1.2 Chapters’ outlines
This section is dedicated to review the main argumentations of the present work in order to provide an overview disaggregated by chapters and their interconnectivity. In Chapter 2, the main characteristics and shortcoming of the New Economic Geography are investigated in the light of its legacy of neoclassical approaches and critics moved by the Evolutionary Economic Geography. In particular, the underestimation of technological externalities and the omitted inter-industry knowledge spillovers are especially underlined. This becomes particularly relevant considering the increasing of knowledge-based economies around the world (Hanusch & Pyka, 2007; Hudson, 2001, 2005; OECD, 1996), as well as in Indonesia (Menkhoff et al., 2011), where innovation emerges as a major competitive driver for firms’ profitability and survive. Following this in Chapter 3, the economic roles play by technological externalities are especially investigated focusing on the theoretical and empirical contributions of economic variety sectoral decomposition. This is conducted analysing first the limitations of highly specialized locations, and then the novelty of (un)related variety conceptualization useful to reduce industry-specific effects. It will be argued that discovering and promoting local relatedness can reduce the risk of lock-in and economic instability. In addition, determining local industrial heterogeneous degree also allows policymakers to develop initiatives towards embedded relatedness and/or further diversification. In Chapter 4, selected discrete-space indices are proposed to measure agglomeration externalities such as specialization, competition, and general variety, where this latter is decomposed into unrelated and related varieties using entropy formula as proposed by Frenken et al. (2007). These indicators will be employed in the empirical analysis in order to unfold their influence on sectoral and firms’ manufacturing expansion within cities and regencies. In addition, discrete-space indicators will be also combined with selected continuous-space indices such as the global Moran’s I, the Moran scatterplots and the LISA statistics in order to identify spatial patterns more accurately of large and medium manufacturing within and across Indonesian locations.
In Chapter 5, Indonesian economy and manufacturing evolutions are explored investigating the diverse policies during different country’s phases. In particular, two main points are highlighted in order to address the present study. First, the Asian Financial Crisis had a substantial negative impact on Indonesian economy and manufacturing highlighting the country’s weaknesses to external shocks. Second, the slow pace of economic recovery associated with the decline of manufacturing encouraged policymakers to develop more sophisticated initiatives based on location and cluster approaches focusing on critical issues for manufacturing growth such as innovation, human capital and spatial inequalities. Although, Indonesian Government is greatly engaged to pursue economic and manufacturing growth, it emerges that more efforts are required in order to enhance innovative environment, qualified job creation, and the localization of technologically advanced industries. Currently, few labour- intensive industries represent the large majority of manufacturing configuration in the country restraining knowledge spillovers, human capital formation, and economic resilience. The lack of manufacturing diversification emerged as a structural issue and a reduction of labour-intensive industries competitiveness undermined the overall manufacturing growth favouring industrial composition change towards higher degree of technology intensity industries. In Chapter 6, the rise and fall of large and medium manufacturing sectors are investigated between 2000 and 2009. Low technology intensity industries substantially decreased their importance, and higher technology intensity sectors increased their manufacturing contributions. Supporting this transformation towards more knowledge-based productions is highly recommended in Indonesia, which increases industrial diversification, balanced growth, productivity, innovation, and the formation of human capital. However currently, manufacturing growth in Indonesia cannot be achieved without revitalizing labour-intensive industries due to their predominant localization. Thus, it is also advisable to develop public policies underpinning the competitiveness of tradition sectors. In addition, agents’ localization differences on average between cities and regencies are investigated, which generate performance differentials of their localized economic activities via their diverse magnitude of agglomeration externalities. This highlights the necessity to empirically discriminate these two diverse types of administrative units in order to avoid spurious inference analysing indiscriminately the entire country.
In Chapter 7, the influence of agglomeration externalities is investigated on average annual employment, value added and labour productivity growth of established sectors and firms between 2000 and 2009. Population density and human capital show antithetic effects between cities and regencies, which can be due to their diverse urbanization tendencies and industrial compositions. Specialized clusters are negatively associated with sectoral and firms’ growth with particular regard to regencies, which are highly specialized. As a result, established manufacturing activities benefit from an increase of heterogeneous industrial configuration within regencies, which reduces industry-specific negative effects. The preponderant role of manufacturing relatedness in general, and in particular of high and medium-high technology intensity related industries emerged indiscriminately by locations. In Chapter 8, the influence of agglomeration externalities is further investigated on the overall manufacturing structure for five-digit sectors and firms during 2000 and 2009. This is extended disaggregating it by technology intensity degrees and two-digit sectors. Specialization emerged as a preponderant source for the overall manufacturing development. Related variety computed based on the Indonesian industrial classification is beneficial for sectoral industrial structure growth with particular regard to cities. A divergent impact of high and medium- high, and medium-low and low technology intensity related industries emerged. Disaggregating manufacturing structure based on technology intensity industries and two-digit sectors, it is observed that economic activities take advantage due to an increase of their technological relatedness. Unrelated variety shows an opposite effect between cities and regencies due to their diverse level of economic density. In this context, human capital comes to light as a predominant driver for manufacturing revitalization regardless to the type of locations and sectors.
In Chapter 9, spatial inequality is investigated highlighting the persistent present of two agglomeration bells around large economic centres in Java. The notion of specialisation and relatedness is combined through the identification of key embedded specialised clusters. Since specialisation and relatedness can be seen as complementary sources to enhance localisation externalities as argued by Jacobs (1969), and the competitive advantages of clusters as supported by Porter (1990), which can reduce the risk of lock-in and local resilience. Thus, policymakers should combine the notion of specialisation and relatedness in a unified policy framework to design more effective public initiatives. The industrial configuration of Eastern Jakarta is used as a case study to unfold the role of key embedded specialised clusters on growth useful to design effective policy for its future industrial development. It is also argued that the current industrial changes towards knowledge-based productions within large economic centres can lead to manufacturing transformation and revitalisation in the country. Since the substantial development of high and medium-high technology intensity industries in Eastern Jakarta (among other dense places) can affect the industrial development across locations, with particular reference to the two agglomeration bells in Java generating spatial snowball effects. Finally in Chapter 10, empirical outcomes are reconciled in the light of their policy implications aiming to revitalise manufacturing in Indonesia. It emerged that Indonesian policymakers should address the following initiatives. 1) Supporting key embedded specialized clusters favouring inter and intra-industry knowledge spillovers. 2) Encouraging population growth and industrial diversity within regencies reducing the negative impact of industry-specific effects; and discouraging them within cities decreasing the risk of over congestion. 3) Underpinning the development of human capital, and the genesis and growth of technologically advanced industries, which can increase manufacturing resilience, further formation of skilled workers, and innovation capabilities, which can be also beneficial for their unrelated activities. 4) Enhancing domestic and international competitiveness of Indonesian manufacturing through favouring sectoral rivalry, which lead to selection of firms making their aggregation more efficient and productive. 5) Developing regional policies in Java Island exploiting spatial industrial development across locations, especially promoting growth of high and medium-high technology intensity industries, which can lead to manufacturing transformation and development across locations.
In Figure 1.1, the structure and key concepts of the present study disaggregated by chapters are illustrated.
Abbildung in dieser Leseprobe nicht enthalten
Figure 1.1: Structure and key contents of the present study.
Notes: The signs between brackets reported for Chapter 7 and Chapter 8 denote the observed influence of the relative agglomeration sources on the explained variable.
2. New Economic Geography’s peculiarities and limitations
2.1 Introduction
For several centuries, there has been a tendency towards urban concentration across the world due to migratory flows from rural to urban areas. Urbanization consistently grew between 1950 and 2011, and it will continue to do so in all regions (United Nations, 2012). For the first time in history more people around the world live in urban centres than in rural areas. In 2011 from Table 1 in the report of United Nations (2012), 3.63 of 6.97 billion people lived in urban areas (slightly above 52%) and by 2050 they project 6.25 of 9.31 billion (slightly above 67%), so population will grow by around 34% and urbanization by 72%. Different nations have experienced urbanization process at different times; developed countries had a faster urbanization before 1950, and developing nations after this date (United Nations, 2012). In 2011 inhabitants living in urban centres in developed regions accounted for around 80% of the population, whereas, Asia and Africa are expected to reach the point where half of their population lives in urban areas by 2020 and 2035 respectively (United Nations, 2012). This tendency of concentration of inhabitants within urban centres can be associated with the maximization of their socio-economic utilities.
Empirical evidence (see, for instance, UN-HABITAT, 2010) demonstrates that there is a significant correlation between urbanization and economic development since GDP per capita, firm and workers concentration tend to increase simultaneously in more urbanized countries, regions and cities. China is a recent example of these linkages, where increased urbanization has fostered socio-economic conditions contributing to poverty reduction as well as improved welfare and an improved standard of living (UN-HABITAT, 2012). The concentration of agents in urban areas may create integrated urban regions forming clusters of cities, for instance, around one or more hub such as Metro Manila, Jakarta, Delhi, or Karachi (Laquian, 2005); or alternatively in the absence of a major hub city, where large and medium-sized cities form an integrated urban region such as in Guangzhou, Shenzhen, Hong Kong and Macau within the People’s Republic of China (Yeh, Yok-shiu, Tunney, & Nien, 2002).
Urbanization and industrialization are unavoidable consequences of the development process of nations, regions, and cities, though different locations make the urban transition at different stages of their country’s development, and with various urban and economic growth patterns (Wheaton & Shishido, 1981; Williamson, 1965). It is, however, possible to find countries where high levels of urbanization fail to generate urban economic development, such as in African nations, including: Madagascar, Niger, Senegal and Zambia, among others (Kessides, 2006). In these cases where urbanization and development have not gone hand in hand, dense urban concentration has generated high unemployment rates, congestion, poverty, low welfare, and poor infrastructures, among other negative consequences (see, for instance, Boadi, Kuitunen, Raheem, & Hanninen, 2005; Fay & Opal, 2000). In particular, local authorities have failed to address suitable policies to support urban concentration and industrial growth creating unsustainable urban development (Boadi et al., 2005).
Given the tendency of agents to concentrate within certain locations and the importance of space, academic attention to economic geography has been increasing over time. Numerous scholars (see, for instance, Christaller, 1933; Isard & Vietorisz, 1959; Krugman, 1991a; Nelson & Winter, 1982; Von Thünen, 1826; Weber, 1909) have investigated reasons for economic agglomeration, what kind of economic activities are concentrated and where they are located. The current predominant framework within the theoretical and empirical economic geography literatures refer to New Economic Geography (NEG) initiated by Krugman (1991a, 1991c), though recently numerous aspects of the approach has been criticised (see, for instance, Boschma & Frenken, 2006; Garretsen & Martin, 2010; Martin & Sunley, 1996). It has been mainly criticised for its legacy of neoclassical approaches and the way to treat technological externalities as a secondary dynamic, where the only Marshallian externality has been considered omitting the important driver of knowledge spillovers across sectors. In order to overcome this limitations, a new conceptualization emerged denominated Evolutionary Economic Geography (EEG), which can be tracked on the seminal contribution by Nelson and Winter (1982). Afterwards, numerous evolutionary studies of economic geography has been elaborated (see, for instance, Boschma & Lambooy, 1999; Boschma & Wenting, 2007; Brenner, 2004; Essletzbichler & Rigby, 2005; Klepper, 2002; Rigby & Essletzbichler, 1997; Swann & Prevezer, 1996), though EEG framework is still under development (Martin, 2003). However, there is no doubt that NEG greatly influenced the agglomeration theories and brought new insights to the study of economic geography. In particular, Krugman (1991a, 1991c) had the merit to combine transportation costs, increasing firms return to scale, and imperfect competition within the full general equilibrium where supply and demand are endogenized (Garretsen & Martin, 2010).
In the study of economic agglomeration, it is essential to determine under which conditions an agglomeration site formed, why certain places grow faster than others, and what factors determine the dispersion of economic activities. This can be explained by the uneven distribution of the “first nature” and “second nature” as argued by Cronon (1991). The former refers to natural endowments such as climate, topography, raw materials, and communication ways, among other factors; and the latter represents the outcomes generated by human behaviours. NEG and EEG substantially differ on the determinants of agglomeration generation and growth. NEG considers pecuniary externalities as the main agglomeration drivers and technological externalities as a secondary dynamic. EEG assumes that technological externalities are the preponderant sources of economic concentration where pecuniary externalities arise as a secondary dynamic due to an increase in competition. Although NEG and EEG consider the initial space as neutral4, they do not dismiss the importance of first nature advantages to explain economic agglomerations, since often clusters are created around natural endowments (e.g. access to the sea, natural resources) to reduce firms’ transportation costs and exploit pre-existing sources.
Natural endowments are distributed unevenly among places generating irregular spatial distribution explaining why certain industries cluster in specific places (Cronon, 1991; Ottaviano & Thisse, 2004). Examples of the influence of natural endowments are the localization of wine producers in California, France, Italy and Australia, and the steel industry near the Great Lakes region in the USA with easy access to iron ore and coal. Ellison and Glaeser (1999) argue that the presence of natural competitive advantages within a location can explain half of geographical colocalization. The mobility of workers and growth of the city can be connected to first nature advantages as argued by Black and Henderson (2003), where natural communications (e.g. ocean) and produced communications (e.g. railroad) play a paramount role on it facilitating the flow of trade with other locations as stated by Beeson, DeJong, and Troesken (2001). Roos (2005), and Rosenthal and Strange (2001, 2004) find evidence that first nature and agglomeration economies are both determinant sources for economic clustering. Since economic agglomeration asymmetry is not determined solely by a sites’ first nature characteristics as many clusters are less natural resource dependent such as Chicago, which became the central city of the America heartland without any natural competitive advantage (Cronon, 1991; Krugman, 1993). Thus, the second nature needs to be taken into account in order to explain the formation and development of economic agglomerations, which modifies the first nature by a multitude of individual actions. Based on Starrett’s work (1978) on the spatial impossibility theorem5, Fujita (1986) observes that in order to take into account spatial agglomeration formation and growth as an endogenous phenomenon three characteristics are fundamental: externalities from non-market interactions made by agents (technological externalities) where the distance among firms plays a prominent role; imperfect market competition, which is an essential condition for the increasing of firms returns to scale (pecuniary externalities); and the heterogeneity of space (natural endowments), which contributes to explain the formation of the central business district (CBD) in a given location.
This chapter explores the main NEG’s characteristics considering the comments made by critics in order to shed light on its limitations6. Particular emphasis will be placed on the underestimation of technological externalities and the omission of a crucial driver of inter-industry knowledge spillover. In Section 2.2, the main NEG’s features are investigated discussing them based on the recent critics moved by the EEG scholars. In Section 2.3, the determinants of agglomeration formation and development under the NEG framework are discussed. The conclusions are provided in Section 2.4.
2.2 The main characteristics and critics of the New Economic Geography
Von Thünen (1826) introduced the theory of agricultural location, designing a framework to optimize land-use for the maximization of farmers’ net profits and consequently his land rents. This was the first early attempt for the theory of location anticipating other future studies of spatial economics (Samuelson, 1983). He investigated agglomeration and dispersion forces, which force individuals and economics activities to move in or out certain places. The pioneering work of Von Thünen (1826) anticipated the Marshallian forces (1890) adopted by Krugman (1991a, 1991c) such as the market-size effect and thick labour market. Von Thünen investigated pecuniary externalities as drivers of agglomerations thought he did not considered pure external (dis)economies. Pecuniary externalities are external to firms’ production activities and generated by product market interactions mediated by the price mechanism (Scitovsky, 1954). They are associated with increasing return at the firm level under imperfect competition where a decision by an agent affects the market price and consequently other agents’ decisions (Fujita, Krugman, & Venables, 1999; Scitovsky, 1954). However, neoclassical economists assume pure competition, thus, a constant return to scale is implicitly considered.
Pecuniary externalities are the central forces of the NEG approach in order to explain under neutral space conditions the causes of economic formation and development of nations, regions and cities. The first NEG’s conceptualization was presented in Krugman (1991a, 1991c), which considered the question of how agglomerations are formed and under what conditions they are (un)stable. Krugman pointed out that concentration of firms can take place through the constant interplay of increasing returns to scale at the firm level, transportation costs and factor mobility (Krugman, 1991a, 1991c). The NEG’s formulation and orientation subsequently have been extended by important works of several authors such as Fujita (1988), Krugman (1995), Krugman and Venables (1995) and Venables (1996). NEG greatly influenced agglomeration theories and brought new insights to the study of economic geography although several of NEG’s ingredients are borrowed from the early location and agglomeration theories, which have been unified and reinterpreted in the NEG framework. NEG introduces little that is new in comparison with the past theories though NEG overcomes their limitations (Fujita, 2000, 2011) such as endogenous growth rather than exogenous growth, imperfect market competition rather than perfect market competition, full equilibrium rather than partial equilibrium, and non- monecentric models rather than moncentric models.
Several aspects of NEG framework have been criticised by EEG scholars (see, for instance, Boschma & Frenken, 2006; Garretsen & Martin, 2010; Martin & Sunley, 1996) as follows. 1) The full general equilibrium and the multiple equilibrium do not reflect the dynamic nature of the real economy; 2) the utility maximization and “representative agents” do not consider the context where the decision is made and the diversity of agents; 3) monopolistic and oligopolistic market structures reflect few real market situations; and 4) technological externalities are considered as secondary dynamic and omitting Jacobian externalities (Jacobs, 1969). In the rest of this section the main ingredients of the New Economic Geography are critically investigated in the light of neoclassical theories and these recent critics moved by the Evolutionary Economic Geography, which are schematically synthesized in Figure 2.1.
Imperfect market competition and increasing firms returns to scale. NEG, as explained in Krugman (1991a, 1991c), assumes that firms choose a location within large imperfectly competitive markets in order to increase their returns to scale and minimize their transportation costs. The imperfect market competition is the necessary condition in order to preserve the assumption of increasing firms’ returns to scale. NEG mainly adopts monopolistic and oligopolistic competition (Fujita & Thisse, Fujita & Krugman, 2004; 2002); in contrast with neoclassical approaches, which embrace the perfect market competition and as a consequence constant returns to scale is assumed. However, pure competition is an idealised market, and monopolistic and oligopolistic markets reflect only few real competition structures. By contrast, EEG considers monopolistic competition based on the Schumpeterian notion of “creative destruction”, which reflects more realistic markets in terms of dynamics and structures. EEG assumes that firms’ innovation capabilities are the main cause of increasing returns to scale due to the development of new products and processes, which lead to temporary monopolies (Grossman & Helpman, 1991). However neoclassical theory, NEG and EEG agree that tough competition fosters convergence among economic activities and locations since fierce rivalry erodes firms’ profitability.
Utility maximization by agents and representative agents. Another concept borrowed from neoclassical theories by NEG is that agents seek and choose a given location to maximize their utilities and profits assuming the homogeneity of agents (“representative agents”). Numerous scholars (see, for instance, Amin & Thrift, 2000; Boschma & Frenken, 2006; Garretsen & Martin, 2010; Hanusch & Pyka, 2007; Martin & Sunley, 1996) criticise this approach since it does not deal with space. The assumption of “representative agents” does not take into account the spatial heterogeneity of locations and the geographical diversity of agents’ competencies and capabilities. Indeed, agents’ decisions should be considered bounded in their rationality rather than just maximization of utility a- context, since agents’ decision processes are highly affected by the context where local institutions play an important role on it, as supported by the EEG (see, for instance, Boschma & Frenken, 2006).
“Iceberg” transportation costs. NEG mainly adopts the “iceberg” transportation costs, which was originally introduced by Samuelson (1954). Transportation costs are computed as a constant percentage of the Free-On Board (FOB) price between two locations, and any increase in the price of transported products implies a proportional increase in the shipment costs (Krugman, 1998). In other words, they are calculated as a constant fraction of the value of shipped goods, which increases proportionally with the distance covered. In the early theory, the transportation costs is a critical factor and their computation can undermine the constant elasticity of demand, which is preserved using “iceberg” transportation costs (Krugman, 1998). However, several researchers (see, for instance, Ottaviano, Tabuchi, & Thisse, 2002) have criticized the “iceberg” transportation costs, in particular, they argued that it is inapplicable in many real situations. Thus, alternative ways to calculate shipment costs have been developed within the NEG framework (see, for instance, Ottaviano et al., 2002).
The full general equilibrium through endogenous growth and multiple equilibria. The full general equilibrium model adopted by NEG assumes that all market processes and firms’ returns are generated endogenously fostering external economies of agglomeration (Krugman, 1991a, 1991c). This was an important contribution by the New Economic Geography in comparison with earlier theories, which partially considered the equilibrium of the system (Krugman, 1998). Since they did not include all economic factors endogenously such as geographical distribution of population, demand and supply. NEG assumes a neutral initial stage where the persistent interaction of agglomeration forces generates a core-periphery configuration. Afterwards, the full general equilibrium and multiple equilibria emerge within and between places restoring the symmetric initial condition due to invisible-hand dynamic processes of agents’ localization decisions, which are oriented towards utility and profit maximization.
EEG criticizes the general equilibrium mechanism as a static equilibrium, which does not reflect the dynamic nature of the real economy (Boschma &
Frenken, 2006) since the system is likely to be in a temporary equilibrium and disequilibrium. EEG assumes that the system is out-of-equilibrium embracing the Schumpeterian notion of market competition. Temporary convergence and divergence are generated due to endogenous firms’ innovation behaviours causing the dynamic distribution of economic activities in space and time (Boschma & Martin, 2010). Firms, through innovation, can have disproportional profitability generating uneven distribution of economic activities, whereas the erosion of profits due to increasing of price competition is considered as a second dynamic, which leads to a short-run economic convergence causing smart selection of organizational routines (Boschma & Frenken, 2006). This process of temporary economic balance and imbalance due to firms’ innovation behaviours is considered recursive.
The non-monocentric urban models. NEG adopts the non-monocentric urban models (see, for instance, Fujita & Ogawa, 1982; H Ogawa & Fujita, 1980; H. Ogawa & Fujita, 1989) overcoming the limitations of the monocentric urban models (see, for instance, Alonso, 1964; Mills, 1967, 1972; Muth, 1969) employed by the early agglomeration theories. Alonso (1964) introduced for the first time the monocentric urban model reinterpreting the Thünen’s framework (Fujita, 2000, 2011), and afterwards it has been extended by several authors (Fujita, 1989; Mills, 1967, 1972; Muth, 1969). The monocentric urban model assumes the existence of a unique market in the city, which is considered the central business district (CBD), and all workers live in the surrounding areas supposing to commute to the CBD. This has been criticized for the assumption that the CBD is formed and grown exogenously.
All economic forces need to be considered endogenous in order to explain the genesis and patterns of CBD (Fujita, 2011; Mori, 2006) and to achieve the multiple equilibria between CBDs. In order to overcome this shortcoming, several economists have elaborated non-monocentric urban models, which are built based on a polycentric approach where the formation of the entire local spatial structure of the economy is endogenously determined assuming the market imperfection. The non-monocentric urban model was introduced for the first time by Fujita and Ogawa (1982) and subsequently it has been extended by several scholars (see, for instance, H Ogawa & Fujita, 1980; H. Ogawa & Fujita, 1989). They demonstrate that market interactions alone under imperfect competition can explain the spatial agglomeration of economic activities between CDBs. These conceptualizations along with the notion of non-monocentric urban models have been embraced by NEG as critical elements of its framework.
Economic (de)agglomeration forces. Pecuniary externalities are considered the main determinants of spatial convergence and divergence of economic activities where technological externalities arise as a consequence of market interactions. Krugman (1991a) justifies the limited importance attributed to knowledge spillovers in NEG since they are difficult to measure given their intangible nature. EEG scholars (Boschma & Frenken, 2006; Garretsen & Martin, 2010; Martin & Sunley, 1996) strongly criticise the assumption of secondary dynamic of technological externalities and taking into account the only Marshallian externalities omitting the important driver of knowledge spillovers across sectors. Krugman (1991a, 1991c) identifies several centripetal forces, which favour concentration of agents; and centrifugal forces, which discourage such proximity. Centripetal forces are the typical Marshallian’s sources (Marshall, 1890): market size effects through linkages, thick labour markets and pure external economies. Centrifugal forces are: immobile factors of production (e.g. lands, natural resources, and people in international context), land rents due to high demand and pure external diseconomies (e.g. congestion).
Krugman and Venables (1995) and Venables (1996) argue that vertical linkages between upstream and downstream industries under imperfect competition can have the same agglomeration role as the migratory inflow of workers á la Krugman (1991a, 1991c). If industries are vertically connected within input-output configuration, downstream markets determine the market size for intermediate products, shaping the size of upstream industries. Many downstream firms generate a large market for intermediate goods (backward linkages) favouring suppliers’ localization. A large upstream market allows downstream industries to have lower transportation and inputs costs (forward linkages) leading to further delocalization of firms. Within a dense proximity of industries, firms would pay higher salaries due to competition for labour and this leads to further workers immigration due to wage differentials between locations. If the increase of firms’ returns to scale within a large market supports higher wages, firms are still encouraged to drive their business within the location; otherwise, dispersion of economic activities is favoured towards other sites with lower salaries and higher firms’ return to scale7.
Circular cumulative causation model. In order to connect and describe the persistent interaction of pro-concentration and anti-concentration forces in a path-dependence way, NEG embraces the circular cumulative causation model. It was introduced for the first time in Myrdal (1957), while Hirschman (1958) included the backward and forward linkages within the model. Afterwards, the circular cumulative causation model has been adapted and applied to a variety of academic fields. It is a multi-causal approach and the idea underling it is that the persistent and accumulative variations of forces produce several changes in the environment. NEG assumes that economic localization is favoured if concentration sources are stronger than dispersion forces within a location; otherwise deconcentration sources force footloose firms and dwellers towards other places considered more economically attractive. The persistent interplay of pro-concentration and anti-concentration forces can generate a threshold-effect of economic agglomeration when a critical level is overcame generating spatial economic balance or imbalance (see, for instance, Durlauf & Johnson, 1995).
The capital accumulation in a given location fosters external economies and historical accidents self-enforcing agents’ expectations, though this latter may arise in the absence of past accidents (Krugman, 1991b). The cause and effect relationship between past and future events may create convergence of agents’ expectations, which lead to economic agglomeration or dispersion. Self-fulfilling and overlapping expectations occur when agents move in or out of a particular place based on their positive or negative expectations that a specific event will take place (Baldwin, Forslid, Martin, Ottaviano, & Robertnicoud, 2003) such as the level of rents and market expansion. The strength of agglomeration economies and convergence of agents’ expectations might also lead to location hysteresis, which is related to a shock in the region and this might cause a catastrophic agglomeration (Baldwin et al., 2003). Temporary shocks might lead to permanent changes in the agglomeration landscape, which might be not reversible. This could be conducted to the effect of the Asian Financial Crisis (1997-1998) in Indonesia, which hardly hits its economy leading to the decline of manufacturing agglomeration causing its composition transformation. The changes in economic and competitive paradigms due to a two-year shock generated selection of manufacturing activities. Since the less competitive and innovative firms and sectors are pushed out from the market or substantially reduced their economic contributions, whereas the more competitive and technologically advanced ones survived and evolved (see Chapter 5 and Chapter 6). The mutation of agglomeration landscape encouraged Indonesian policymakers to underpin the new manufacturing growth pathway where human capital and knowledge spillovers emerge as pillar factors in leading manufacturing revitalization and its transformation in the country (see Chapter 7 and Chapter 8).
2.3 Agglomeration formation and development under NEG
Urbanization economies were introduced by Hoover (1937) discerning them from localization externalities. The former is internal to the city and external to the industry fostering the output of all firms within a location, which increase the dimensions of the overall economy. The latter is internal to a given industry and external to the firms increasing the outputs of localized economic activities with the same industry. As argued by numerous scholars (see, for instance, Frenken, van Oort, & Verburg, 2007; Harrison, Kelley, & Gant, 1997; Henderson, 1986; Van Oort, Burger, Knoben, & Raspe, 2012), urbanization externalities are more associated with the local demand effect á la Krugman (1991a, 1991c). However, Henderson (1986) argues that local demand does not explain fully why firms from different industries want to locate in close proximity to each other underpinning the preponderant role of knowledge spillovers within the same sector.
Abbildung in dieser Leseprobe nicht enthalten
Figure 2.1: The main NEG’s peculiarities and critics.
Krugman’s framework (1991a, 1991c) begins with the migration inflows of workers due to wage differentials between locations, albeit workers are less mobile in the international context due to the difficulty for companies to recruit them. The inflows of labour enlarge local demand increasing firms return to scale (backward linkages, demand side). The market size effect through linkages generates a “snowball” mechanism increasing labour market pool, and the concentration of downstream firms, which increase the demand size of intermediate products and services. As a result, upstream industries are encouraged to move into the location fostering input-output vertical linkages with positive implications on firms’ returns to scale, and intra-industry knowledge spillovers emerge as a secondary dynamic. The proximity of agents decreases input and transportation costs, and market prices (forward linkages, cost side). This increases productivity and profitability with positive repercussion to nominal wages. The rise in workers’ salaries and lower product prices increase real wages supporting firms’ productions of diversified goods in order to satisfy large heterogeneity customers’ needs. Product differentiation fosters further economic agglomeration as enterprises can avoid price competition (Fujita & Thisse, 2002), and workers are encouraged to be in the place with availability of jobs, high salaries, reduced market prices and large product varieties. As aforementioned, input-output configuration can have the same agglomeration role as the migratory inflow of workers (Krugman & Venables, 1995; Venables, 1996), which also arises knowledge spillovers, albeit NEG considers only the Marshallian externalities (Marshall, 1890) neglecting the role of Jacobian externalities (Jacobs, 1969).
Fugal sources make locations less competitive in attracting agents (e.g. high levels of rents and local congestion) generating a negative path-dependence mechanism. When a location becomes densely concentrated, factor market competition and product market competition arise the negative forward linkages (Fujita & Thisse, 2002; Krugman, 1991a, 1991c; Puga & Venables, 1998). Firms are encouraged to be in a location until when the benefits related to the increasing firms’ returns to scale overcome the drawbacks related to the raise of nominal wages and the overall production costs. Agglomeration and dispersion forces are connected through the circular cumulative causation, which highlights the possibility of forecasting a given event raising agents’ expectations convergence. They can play as well as a role in self-enforce economic agglomeration or dispersion fulfilling or overlapping a particular location based on agents’ prediction (Baldwin et al., 2003). The interaction between agglomeration forces and agents’ expectations might also generate location hysteresis, which might lead to catastrophic agglomeration (Baldwin et al., 2003). Figure 2.2 synthesizes the path-dependence mechanism of these agglomeration forces.
Abbildung in dieser Leseprobe nicht enthalten
Figure 2.2: The main NEG’s agglomeration forces linked through the circular cumulative causation.
NEG assumes a core-periphery configuration in the first stage due to the persistent interaction of agglomeration forces, and then a mechanism of self- organization is generated restoring full general equilibrium within the system and multiple equilibria among locations. An evident example of the core- periphery configuration can be found in Italy, where the historical internal mobility from Southern to Centre-northern regions has generated a large socio- economic asymmetry within the country (Piras, 2012). The migratory flows have fostered the growth of the Centre-northern area, what has become known as the “Third Italy”, generating a core-periphery configuration in a dualistic relationship “North-South”. Although this phenomenon began on the end of the War World II, significant socio-economic differences are still in place between these two macro Italian regions. Therefore, the mechanism of self-organization as theorized by Krugman (1991a, 1991c) did not occur in Italy. This conceptualization is under academic debate since it does not reflect the dynamic nature of the real economy (Boschma & Frenken, 2006; Boschma & Martin, 2010).
2.4 Conclusions
In this chapter, the main ingredients and limitations of the New Economic Geography have been investigated in the light of its legacy of neoclassical approaches and critics moved by Evolutionary Economic Geography. It is no in doubt that NEG greatly influenced the agglomeration theories and brought new insights within the economic geography; though important limitations have been highlighted, therefore, a question emerges: Do we need to rethink or overcome NEG framework? This question is currently under academic debate; this chapter suggested that NEG framework should be reconsidered in order to explain consistently and coherently the varieties of agglomerations genesis and development. In particular considering the increasing importance of knowledge- based economies around the world (Hanusch & Pyka, 2007; Hudson, 2001, 2005; OECD, 1996), where innovation emerges as a major competitive driver for firms’ profitability. In this context, EEG can represent a potential alternative, evolutionary studies are taking ground among researchers though its theoretical framework stills under development.
This chapter highlighted an important drawback of NEG framework referring to the underestimation of technological externalities, which are considered as a secondary dynamic and omitting knowledge spillovers across sectors. The following chapter embraces the conceptualization that technological externalities are the pillar determinants for economic growth in the light of the recent contribution of economic varieties decomposition proposed by Frenken et al. (2007), which provides new theoretical and policy insights for researchers and policymakers.
3. The role of agglomeration externalities on economic growth
3.1 Introduction
In the previous chapter, the limitations of the New Economic Geography have been highlighted. In particular, it has been argued that Krugman (1991a, 1991c), in his seminal works, attributed little importance to technological externalities and without taking into account inter-industry knowledge transmission missing an important building block within the puzzle of regional and urban development. This assumes important connotations considering the recent increase of knowledge-based economies around the world (Hanusch & Pyka, 2007; Hudson, 2001, 2005; OECD, 1996), which arises the necessity to take into account innovation as a crucial driver for economic growth of nations, regions, sectors, and firms . The present study embraces this notion where knowledge spillovers stand behind the generation of innovation, which allows firms to have disproportional profitability, as argued by EEG scholars.
Although, there is a general agreement among researchers that knowledge generation and spill over play an important role in regional innovation and growth (Karlsson & Manduchi, 2001), for more than two decades, scholars debate on the following matters. If the creation and diffusion of knowledge between actors is a function of distance (Bathelt, Malmberg, & Maskell, 2004; Rallet & Torre, 1999), which technological externality is more important for location growth and under which market structure innovation is optimized (see, for instance, Beaudry & Schiffauerova, 2009; De Groot, Poot, & Smit, 2009; Feldman & Audretsch, 1999; Van der Panne, 2004). Which type of industries are more responsive to which externality (see, for instance, Boschma & Frenken, 2009; Neffke, Henning, Boschma, Lundquist, & Olander, 2011). Several authors (De Groot et al., 2009; Puga, 2010; Rosenthal & Strange, 2004; Van Oort, 2007) also argue that the effects of agglomeration economies differ across sectors, space and time. This debate increased over time since researchers have found evidence to support different theoretical conceptualizations (see, for instance, De Groot et al., 2009; De Groot, Poot, & Smit, 2015). A potential source of this inconclusive debate may stem from the diverse types of sectors and level of analysis, different stage of industry life cycle examined (Bishop & Gripaios, 2010; Neffke, Henning, Boschma, et al., 2011; Paci & Usai, 1999; Van Oort, 2004), methodologies employed and the misspecification of economic variety (Boschma, Minondo, & Navarro, 2012). With regard to this latter, economic variety decomposition based on sectoral linkages addresses the misspecification of Jacobian externalities, which can contribute to resolve the aforementioned long-term academic debate.
Agglomeration economies can be categorized into four main forces explaining knowledge spillovers and economic agglomerations formation and evolution in different ways as follows. 1) Firms are encouraged to operate in proximity within the same industry due to intra-industry knowledge spillovers (Glaeser, Kallal, Scheinkman, & Shleifer, 1992). 2) Firms take advantage of locating their activities close to complementary industries exploiting inter-industry knowledge spillovers (Jacobs, 1969). 3) Economic localization occurs regardless of the nature of established industries since benefits arise from a dense and heterogeneous environment within a location, in terms of population, R&D centres and business services among other “pull” forces, which foster the outputs of all firms localized in the area (Hoover, 1937)8. 4) Knowledge transmission and economic growth is influenced by the degree of competition (Glaeser et al., 1992; Jacobs, 1969; Porter, 1990), which can be also associated with the notion of Darwinian selection and adaptation of ecologic system (Combes, Duranton, Gobillon, Puga, & Roux, 2012; Duranton & Puga, 2003; Melitz, 2003; Melitz & Ottaviano, 2008). Agglomeration externalities can be generated through industrial configuration of inter and intra- industry establishments and/or large market size generating a mechanism of economic path-dependency.
There is no doubt that economic proximity arises agents’ benefits, which can be associated with sharing facilities and infrastructures, availability of a large and skilled labour pool, large and heterogeneous suppliers, gaining from external economies, better matching between agents, and learning through knowledge exchange due to interactions between agents (Duranton & Puga, 2003). However, agents’ concentration increases agglomeration costs such as pollution and congestion, among others. The trade-off between agglomeration benefits and costs makes a location more or less competitive in attracting economic activities and dwellers. A large body of theoretical and empirical literature has been made in order to investigate why firms and workers prefer to be within highly concentrated places albeit this increases their costs (see, for instance, Melo, Graham, & Noland, 2009; Puga, 2010; Rosenthal & Strange, 2004). Empirical evidence shows that firms and workers have higher performance within a large and dense economic environment and this can be associated with the proximity effect of economic activities, from which arises agglomeration externalities (see, for instance, Duranton & Puga, 2003; Melo et al., 2009; Puga, 2010; Rosenthal & Strange, 2004).
This chapter is devoted to establish a conceptual relationship between the types of agglomeration externalities and their economic roles in the light of recent
contribution by Frenken et al. (2007), which distinguishes variety based on sectoral linkages into related and unrelated varieties. Four main contributions are identified stemming from this decomposition as follows. First, the idiosyncratic role played by inter-industry knowledge spillover (Jacobs, 1969) and portfolio diversification (Conroy, 1974, 1975) effects can be separately evaluated addressing the misspecification of Jacobian externalities. Second, (un)linked variety and urbanization externalities can be conducted to more appropriate theoretical foundations; where Jacobian externalities are associated with the role played by related variety, urbanization externalities are linked to the market-size effect through linkages á la Krugman (1991a, 1991c), and unrelated variety is connected to the portfolio diversification. Third, discovering economic relatedness allows policymakers to develop ad hoc initiatives to promote key industries with large intersectoral linkages, which permits to reduce the risk of lock-in effect and lack of resilience (typical drawbacks of having a location highly specialized). Fourth, the identification of local economic embeddedness and heterogeneous configuration degrees allows more accurate policy strategies to examine and pursue economic growth and diversification.
The rest of this chapter is organized as follows. In Section 3.2, the economic role of technological externalities are explored in the light of decomposition of economic varieties based on linked sectors distinguishing related and unrelated varieties. This brings new important implications for policy design, which are also investigated. In Section 3.3, the effect of competition externalities on knowledge spillovers is explored, which can be also associated with smart selection and adaptation of ecologic system. In Section 3.4, the dynamic mechanism of economic development is proposed highlighting the inter-relation of agglomeration externalities, which shape agents’ proximity configuration within locations. Finally, conclusions are provided in Section 3.5.
3.2 Technological externalities
Technological externalities are associated with nonmarket interactions where the activity of a single firm directly affects the production function of other economic agents (Scitovsky, 1954). The first conceptualization of technological externalities goes back to Marshall (1890), who argues that knowledge is a production input diffused freely in the atmosphere due to the dynamic interactions between economic agents within specialized clusters, and it does not require market mechanisms to make it available to users. Afterwards, this conceptualization has been extended by numerous authors (see, for instance, Glaeser et al., 1992; Jacobs, 1969). Neoclassical economists assume that knowledge is a public good, and thus, it is not profitable; though the taxonomy of knowledge is complex by nature, often local and tacit, which is not available to all agents and it does not occur automatically (“in the air”) (Breschi & Lissoni, 2003; Capello, 1999). However, knowledge can be explicit (documented and codified, such as patent documents, scientific and technical literatures), which can be transferred easily to others. Thus, some aspects of public good can be also found since more than one firm can use an idea at the same time (non-rivalry) and it is difficult to exclude specific firms to exploit it (non-excludability).
Knowledge creation and spill over stand behind innovation, which is not only generated at the firm level but often at the meso level through sectoral linkages. Know-how transmission can occur in many different ways such as imitations, spin-off, social networks, labour mobility, and collaborative networks (Boschma & Frenken, 2006). An essential condition of knowledge flow is the dissimilarity of agents’ know-how otherwise lock-in effect can be generated where the distance still plays an important role on it. Since ideas are easier to be transfer among firms in proximity rather than far way (Jaffe et al., 1993), though this can occur between economic activities detached from the regional context due to recent technological progress (Breschi & Lissoni, 2001). Schumpeter (1912, 1942) is one of the first scholar to stress the importance of innovation for economic growth. The Schumpeterian growth model incorporates the technological process as an endogenous introduction of product and/or process innovations by economic agents in order to maximize their utilities and profits. This is the result of the persistent and dynamic interactions between economic activities enhancing their competences and capabilities to innovate through knowledge exchange. This generates a temporary firms’ disproportional profitability self-enforcing location attractiveness of new entrants increasing the diversification of knowledge.
This section aims to investigate this crucial role of technological externalities for economic growth in the light of the contribution of economic varieties decomposition, which provides new insights for researchers and policymakers. This is addressed as follows. In Section 3.2.1, Marshall-Arrow-Romer (MAR) model (Glaeser et al., 1992) is explored highlighting the limitations of highly specialized locations. In Section 3.2.2, the contribution of economic varieties decomposition is examined investigating separetely Jacobian externalities (Jacobs, 1969) and portfolio diversification (Conroy, 1974, 1975) effects, and new relevant insights for tailor-made policies are also discussed.
3.2.1 Knowledge exchange within specialized clusters and its shortcomings
Marshall (1890) examined pecuniary and technological externalities in order to explain the formation and development of economic agglomerations, and he theorized the concept of external economies in the production process within specialized clusters. Marshall (1890) argued that agglomeration externalities encourage firms to produce in proximity to other enterprises within the same industry. Since a specialized economic cluster allows enhancing network of relationships, firms’ innovation capabilities, labour pool and specialized workers, and reducing agents’ transaction and coordination costs. Afterwards, Glaeser et al. (1992) formalized and extended the Marshallian externalities
3.2.2. The contributions and challenges of economic variety decomposition
combining the works of Arrow (1962) and Romer (1986), into what has become known as the Marshall-Arrow-Romer (MAR) model. The MAR model assumes that knowledge spillovers are predominantly industry-specific as intra-industry linkages foster innovation and growth within locations. There are numerous empirical examples of industrial specialization, for instance, the software industry in California’s Silicon Valley in the United States and Bangalore in India, automotive manufacturing in Detroit in the United States, biotechnology industry in Cambridge in the United Kingdom, and the ceramic tile and textile manufacturing in Sassuolo and Prato respectively in Italy.
Although it is expected higher economic performance within specialized places and this has been supported by numerous empirical evidence (see, for instance, De Groot et al., 2009, 2015), two important drawbacks are associated with highly specialized locations: lock-in effect and lack of economic resilience. Lock- in effect can be generated in the long run due to the reduction of know- how complementarity within the same industry. Knowledge transfer over time increases the cognitive proximity between firms reducing their diverse expertise causing a less effective learning process (Boschma, 2005; Nooteboom, 2000). However, the presence of strong knowledge bases and tight external linkages within an industrial cluster allow to overcome the risks associated with lock-in effect, since new external knowledge can spill over within a specialized cluster (Giuliani & Bell, 2005; Graf, 2011), what is called knowledge gatekeepers.
With regard to economic resilience, a location characterised by a high level of specialization is less protected to external industry-specific demand and supply shocks, and technological paradigm shifting due to a lack of portfolio diversification. As argued by Porter (1990), institutional organizations need to create the environmental conditions necessary to sustain the genesis and development of diversified agglomerations, since the future success of a cluster is unpredictable. Even, Marshall (1890) did not dismiss the benefits for a location from having some degrees of industrial diversification in order to increase its economic resilience. It will be argued that the identification of relatedness allows reducing these risks associated with highly specialized locations by promoting key clusters characterized by large sectoral interconnectedness.
3.2.2 The contributions and challenges of economic variety decomposition
Jacobian externalities are commonly measured as general variety without differentiating sectoral linkages though it incorporates two idiosyncratic economic effects: inter-industry knowledge spillovers (Jacobs, 1969) and portfolio diversification (Conroy, 1974, 1975). Recently, Frenken et al. (2007) suggest disaggregating general variety into related and unrelated varieties based on sectoral interconnectedness in order to measure more accurately their idiosyncratic economic roles. This decomposition stems from the conceptualization that related variety is more associated with the role of Jacobian externalities and unrelated variety is more linked to portfolio diversification effect. In the rest of this section, these two effects are separately investigated highlighting their implications for policy design.
3.2.2.1 Knowledge transmission across sectors
In contrast with the MAR model, Jacobs (1969) argued that the creation and diffusion of knowledge are more relevant between complementary industries rather than within the same industry, since innovation generated by an industry can be applied to other related industries. This also drives localization economies. Knowledge is expected to spill over between related industries with some degree of cognitive proximity rather than unrelated industries with large degrees of cognitive distance (Boschma & Iammarino, 2009; Frenken et al., 2007; Nooteboom, 2000). However when the cognitive proximity is too high among agents (specialization) can generate lock-in effect as the relevance of the learning process becomes less effective due to the similarity of agents’ expertise (Boschma, 2005; Nooteboom, 2000). Porter (1990, 2003) is one of the first scholar to recognize the importance of related industries to enhance the competitive advantage of clusters.
Following the recent work of Frenken et al. (2007) in the Netherlands, several empirical studies have been conducted investigating the role played by related and unrelated varieties on innovation, employment and productivity growth in developed economies. Bishop and Gripaios (2010) in Great Britain, Boschma and Iammarino (2009) and Quatraro (2010) in Italy, Boschma et al. (2012) in Spain, Quatraro (2011) in France, Hartog, Boschma, and Sotarauta (2012) in Finland, and Castaldi, Frenken, and Los (2014) in US. These scholars found evidence that related variety fosters regional expansion albeit their approaches widely vary in terms of, for instance, geographical scales, measures of relatedness, periods covered, control variables employed and the country of analysis. In addition, several authors (Boschma, Minondo, & Navarro, 2013; Hidalgo, Klinger, Barabasi, & Hausmann, 2007; Neffke, Henning, & Boschma, 2011) have demonstrated that related variety generates incremental and radical innovations via spin-off, recombination and accumulation of complementary competences, and assets increasing diversification through the creation of regional (un)related branching.
The learning process between interconnected industries is more intense than unrelated activities, which is expected to generate the emergence of new industries and technologies. Knowledge transmission between linked sectors enhances their innovation capabilities favouring the establishments of new relatedness. This can also generate changes, which can be adopted by unrelated industries creating regional unlinked branches guiding to new directions of growth and new market opportunities enhancing local expansion and diversification. However, the genesis of new activities is likely to be related to the
3.2.2.2. Portfolio diversification effect
pre-existing local industrial structure rather than disconnected to the established configuration. Since linkages between economic activities are facilitated and knowledge transmission is favoured, which increase sectors and firms’ survival within a regional embedded space. For instance, Klepper and Simons (2000) use a case-study in the television receiver industry in US demonstrating the creation of new regional related branching. Neffke, Henning, and Boschma (2011) find evidence in Swedish regions that a new industry is likely to establish its activities in a region where other industries are technologically related, and an existing industry is likely to exit in absence of technologically relatedness within a region. Boschma and Wenting (2007), and Klepper (2007) demonstrate that related branching process also increases firms’ survival chances. These studies highlight the positive role of regional related industries on growth and diversification. A well-known example is the case of the Emilia-Romagna region in Italy where the regional engineering knowledge-based favoured the proliferation and expansion of related industries for the production of irradiation, electromedical and electrotherapeutic equipment and luxury car in Modena, manufacture of agricultural and forestry machinery in Reggio nell’Emilia and Modena, and ceramic tile in Sassuolo, among other cognitive proximity clusters (see, for instance, Ercole, 2013).
In addition, the persistent presence of related industries in a location generates regional knowledge-based related-skills, which can contribute to reduce the impact of economic shocks and downturns through absorbing the negative industry-specific fluctuations of demand and/or regenerating the industrial structure into new pathways of growth. For instance, Pittsburgh witnessed a rapid economic recovery due to its strong steelmaking skills supported by related businesses (i.e. steelmaking equipment, engineering services, high-tech devices, and basic refractory brick) (Treado, 2010); and Boston experienced an economic restructuring over the long period of time due to its complementary expertise (Glaeser, 2005). This phenomenon of diversification through economic relatedness recently emerged within the theoretical and empirical literature as a new address of study for local growth and stability. Figure
3.1 schematically illustrates the diversification role of related varieties through knowledge recombination and accumulation between interconnected clusters, which generate new (un)linked branches affecting location resilience and growth.
3.2.2.2 Portfolio diversification effect
The decomposition of economic varieties based on sectoral linkages allows identifying the degree of heterogeneous configuration within a location, which is associated with the portfolio diversification effect. Economic diversity increases location stability protecting from external industry-specific demand and supply shocks, and technological paradigm shifting (Essletzbichler, 2007; Frenken et al., 2007). This also reduces regional economic volatility since a heterogeneous configuration can have a more balanced growth where given
Abbildung in dieser Leseprobe nicht enthalten
Figure 3.1: The diversification process of complementary competences accumulation and recombination into new and related pathways of growth.
sectors perform better than others. The portfolio diversification effect was originally conceptualized and adopted as a strategy to reduce the risk of financial assets through diversification (see, for instance, Markowitz, 1959). Afterwards, Conroy (1974, 1975) suggests a portfolio-theoretic approach to regional economic diversity and diversification in order to reduce the risk of regional instability associated with high degree of specialization in a location. An example is Detroit, which is the most populous city in the state of Michigan and highly specialized in automotive industry. Although this facilitated the city’s growth due to localization economies, Detroit recently experienced an economic downturn due to a significant reduction of global automotive demand, which generated socio-economic instability with unemployment rate of 20% (E. Hill et al., 2012). This was due to the negative industry-specific effect, which could not be absorbed by other industries due to low degree of portfolio diversification within the city.
The relationship between regional stability and performance has been investigated for quite long time by numerous authors (see, for instance, Malizia & Ke, 1993; McLaughlin, 1930; Tress, 1938; Wagner & Deller, 1998), and they find evidence that a location with more economic diversity experienced in more economic stability. The process of economic diversification within a location can be seen as a dynamic mechanism of production, consumption and trade pattern changes (Schuh & Barghouti, 1988) where the degree of heterogeneous
3.2.2.3. Policy implications of economic variety sectoral decomposition
configuration can be associated with the scale and diversity of local demand. However, economic diversity does not mean absence of specialization but the presence of multiple specializations within a location (Malizia & Ke, 1993) where their establishment is not purely random but a certain degree of coherence can exist between related established sectors (Neffke, Henning, & Boschma, 2011). As aforementioned, existing industrial configuration within a location is likely to be related to the past composition generating the future local structure (Neffke, Henning, & Boschma, 2011) where regional interconnected activities increase the probability to survive of new industries and firms (Boschma & Wenting, 2007; Klepper, 2007). Nevertheless, knowledge flow is not precluded for unrelated variety as demonstrated by Castaldi et al. (2014). These scholars investigate the influence of (semi-)related and unrelated varieties on patents in US and they find evidence that the combination of unrelated knowledge can produces radical innovations generating technological “breakthroughs”. This study add a novelty in comparison of the original work of Frenken et al. (2007), which neglects the flow of knowledge between unrelated economic activities, which is rare but it can not be excluded.
3.2.2.3 Policy implications of economic variety sectoral decomposition
There is no doubt that knowledge transmission within a cluster foster firms’ innovation capability and growth, which has been supported by numerous empirical evidence (see, for instance, De Groot et al., 2009, 2015). However, two important constraints of growth emerge within highly specialized locations: lock- in effect and lack of economic resilience. Reconceptualising economic variety based on sectoral linkages can overcome these two drawbacks by discovering and promoting key specialized clusters characterized by large inter-sectoral linkages. New external knowledge can flow between interconnected economic activities with diverse but complementary know-how reducing the risk of similarity of their expertise. The promotion of related variety can also increase location diversification through the formation of regional (un)related branches generating new local growth pathways (Boschma et al., 2013; Hidalgo et al., 2007; Neffke, Henning, & Boschma, 2011).
Policymakers should select and promote a cluster not only based on the value that it can create by itself, but it should be assessed in a broader local prospective based on its contribution to other linked sectors stimulating local growth and diversification. This reduces the risk of a cluster’s failure since related and supportive businesses are crucial elements for the competitiveness of a specialised agglomeration as argued by Porter (1990). However, policymakers should avoid investing in industries that are not actually (or potentially) embedded within their regional context (linked with other sectors); and they should stay away from supporting stagnant and decline clusters (even if they are regional embedded) that show non-temporary competitive weaknesses and/or an enduring reduction of their demand due to their technological paradigms and customers’ preferences changes. The assessment of a cluster’s potentiality represents a policymakers’ challenge since its future success is unpredictable, and the complexity of forecasting is augmented due to the regional embeddedness considerations. This discovery process requires a heedful evaluation of the impact of a cluster on regional structure growth and a careful monitoring during the policies implementation in order to assess their impact on cluster’s evolution and its local contribution.
Policymakers recognize the importance of promoting key related sectors in order to enhance local growth, though the definition and identification of cognitive proximity linkages between sectors, and how the promotion of industries with certain large inter-linkages impacts locations, sectors and firms’ growth represent further policymakers’ challenges in order to develop ad-hoc regional policies. As argued by Siegel, Johnson, and Alwang (1995), the identification of sectoral interconnectedness should be based in terms of explicit economic relationships as type of sector and sectoral interaction based on, for instance, production process and inputs, technology used, and sharing the same infrastructures, among others. Examples of public policies in promoting key industrial clusters can be found in the State of Texas through the Industrial Cluster Initiative, which aims to increase the strength of log-term competitiveness of primarily technology-based industries (Office of the Governor of Texas, 2004). The findings of the Culliton Report are supported in Ireland, which recommend the promotion and the development of clusters and their related industries in order to increase the national competitive advantage in the view of Porter (Doyle & Connell, 2007). From 2004, Indonesian policies (i.e. The National Long Term Development Plan 2005-2025, and the Master Plan for the Acceleration and Expansion of Indonesia’s Economic Development 2011-2025) began to prioritize key industries based on cluster and regional approaches recognizing the importance of local specificity of agents’ localization and agglomeration externalities as contributors to growth (see Chapter 5).
Based on these recent Indonesian policies, it will be argued that decomposing economic varieties based on sectoral linkages can provide valuable insights for policy design in order to revitalize manufacturing activities in Indonesia. This becomes particular relevant considering that its economy progressively moves towards a knowledge-based economy (Menkhoff, Evers, Wah, & Fong, 2011), especially, manufacturing activities witnessed a significant growth of high and medium-high technology intensity industries between 2000 and 2009 (see Chapter 5 and Chapter 6). In this context, learning process plays an increasing role for productivity and employment growth in Indonesia. The identification of relatedness with Indonesian locations allows policymakers to develop ad hoc strategies enhancing knowledge spillover and diversification underpinning manufacturing and location growth. Scholars have commonly focus their attention on the impact of relatedness on regional economic development, and policymakers largely ignore the relationship between growth and stability
[...]
1. ASEAN mainly aims to enhance socio-economic growth and cooperation, regional stability and peace, mutual assistance, educational and research system, and a more effectively utilization of agriculture and industries resources and trades in the region. The country’s members are: Indonesia, Malaysia, Philippines, Singapore, Thailand (co-founders), and Brunei, Cambodia, Laos PDR, Myanmar, and Vietnam.
2. The Indonesian administrative area is divided into provinces, subdivided into regencies and cities, which are further decomposed into districts and then villages. Regencies and cities are at the same administrative level and they have their own local government, legislative body, and a wide autonomy on economic policies following the Indonesian decentralization process initiated by the Law N. 22 and 25/1999, which came into force in 2001, and subsequently amended (for a discussion, see, for instance, Firman, 2009).
3. General variety term refers to Jacobian externalities computed in the old fashion without considering any distinction of sectoral interconnectedness.
4. NEG and EEG assume the neutral space condition since they argue that economic agglomeration and consequently regional development can also occur without any natural endowment differences. However, they differ in the final assumptions, where NEG assumes that the interaction of agglomeration forces restores the symmetric initial condition, and EEG embraces the Schumpeterian notion of temporary convergence and divergence of the system and between places, which is considered recursive.
5. Starrett (1978) refers to spatial impossibility theorem as the incompatible combined notions of agents’ concentration with competitive equilibrium, which is supported by neoclassical economists. Since, interaction among agents generates some kinds of market imperfection making the space inhomogeneous, and agents’ localization decisions are based on geographical differences.
6. This chapter has been constructed in a similar conceptual fashion of several scholars’ works combined (see, for instance, Boschma & Frenken, 2006; Garretsen & Martin, 2010; Martin & Sunley, 1996), which criticize the NEG approach. Part of the present theoretical framework is included in Ercole (2012).
7. Agglomeration forces linked through the path-dependence mechanism as assumed by NEG are synthesise in Figure 2.2.
8. Often, urbanization externalities are improperly associated with Jacobian externalities due to the misspecification of inter-industry knowledge spillovers. The present study clearly distinguishes them, where the former is conducted to the Krugman’s conceptualization referring to Chapter 2, and the latter is linked to the notion of related variety since knowledge is likely to be transmitted between connected activities rather then disconnected ones.
- Arbeit zitieren
- Roberto Ercole (Autor:in), 2017, The Impact of Agglomeration Externalities on Manufacturing Growth within Indonesian Locations, München, GRIN Verlag, https://www.grin.com/document/416775
-
-
-
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen. -
-
Laden Sie Ihre eigenen Arbeiten hoch! Geld verdienen und iPhone X gewinnen.