«TI 2012-072/3 Tinbergen Institute Discussion Paper Production Externalities in the Wood Furniture Industry in Central Java Roos K. Andadari1 Henri ...»
Tinbergen Institute Discussion Paper
Production Externalities in the Wood
Furniture Industry in Central Java
Roos K. Andadari1
Henri L.F. de Groot2
Satya Wacana Christian University;
of Economics and Business Administration, VU University Amsterdam, and
Tinbergen Institute is the graduate school and research institute in economics of Erasmus University
Rotterdam, the University of Amsterdam and VU University Amsterdam.
More TI discussion papers can be downloaded at http://www.tinbergen.nl
Tinbergen Institute has two locations:
Tinbergen Institute Amsterdam Gustav Mahlerplein 117 1082 MS Amsterdam The Netherlands Tel.: +31(0)20 525 1600 Tinbergen Institute Rotterdam Burg. Oudlaan 50 3062 PA Rotterdam The Netherlands Tel.: +31(0)10 408 8900 Fax: +31(0)10 408 9031 Duisenberg school of finance is a collaboration of the Dutch financial sector and universities, with the ambition to support innovative research and offer top quality academic education in core areas of finance.
DSF research papers can be downloaded at: http://www.dsf.nl/ Duisenberg school of finance Gustav Mahlerplein 117 1082 MS Amsterdam The Netherlands Tel.: +31(0)20 525 8579 Production Externalities in the Wood Furniture Industry in Central Java Roos K. Andadaria, Henri L.F. de Grootb,1 and Piet Rietveldb a Satya Wacana Christian University, Faculty of Economics and Business 52-60 Diponegoro Street, Salatiga, Indonesia b VU University Amsterdam, Dept. of Spatial Economics De Boelelaan 1105, 1081 HV Amsterdam, The Netherlands and Tinbergen Institute, Amsterdam-Rotterdam, The Netherlands Abstract This paper exploits micro firm level data to examine the impact of spatial clustering and links to foreign buyer networks on firm performance in the wood furniture industry in Central Java, Indonesia. The analysis is based on an annual manufacturing survey. We identify the impact of specialization of the cluster, diversification, and links to foreign buyer networks. For this purpose, a production function framework is developed. The results lend support to the view that clustering of large and medium scale specialized firms improves firm performance, while clustering of small scale specialized firms and clustering of diverse firms are not conducive to firm performance. We also find a clear positive association between involvement in exporting activities and firm performance.
Keywords: Productivity, Externalities, Wood Furniture Industry, Indonesia JEL codes: D20, R11, R32 Corresponding author: Henri L.F. de Groot, Dept. of Spatial Economics, VU University Amsterdam, De Boelelaan 1105, 1081 HV Amsterdam, The Netherlands, Email: firstname.lastname@example.org. We are grateful to Enno Masurel for comments on earlier versions of this paper. The usual disclaimer applies.
1. Introduction Location externalities, which are derived from industrial clustering, play an important role in fostering competitive advantages and overcoming growth constraints (Schmitz, 1999). Agglomeration theory stresses that the proximity of firms provides benefits, even though the views regarding the sources of the externality benefits are diverse among scholars in this field (see, for example, De Groot et al., 2009, and Melo et al., 2009, for meta-analyses on the empirical evidence of agglomeration externalities). Probably the best-known source of externalities was put forward by Marshall (1920) and was later formalized by Arrow (1962) and Romer (1986). This source is currently known as ‘MAR externalities’ after Alfred Marshall, Kenneth Arrow and Paul Romer. It emphasizes that knowledge accumulation is predominantly an industry-specific process which benefits from specialization and concentration in space. In contrast, Porter (1990) stresses that clusters enhance growth due to competition among specialized firms which stimulates them to innovate. Finally, Jacobs (1969) insists that agglomeration economies are obtained from the proximity of diverse firms.
However, external benefits enjoyed by firms do not come merely from the spatial concentration of firms, but also from other factors such as linkages with external partners. There is some evidence of firms that perform well without being located in a local cluster (e.g., Shaver and Flyer, 2000). Several studies show that firms, especially in developing countries, can gain advantages by establishing a network with external actors, especially international players. The global value chain (GVC) and global production network (GPN) theories (Gereffi, 1999; Ernst and Kim, 2002) emphasize that a firm can benefit from integration in a global buyer value chain or production network. In addition, the foreign direct investment (FDI) theory developed by Dunning (1993) stresses that becoming a partner in the internationalization decisions of international firms may increase firm’s capability by providing access to technology transfer.
Based on the argument that sources of externalities come from location externalities and international network externalities, we propose a simple integrated framework in which externalities cause technological progress and improved firm performance. The changes in technological capabilities can be analyzed in a simple production function framework. The production function reflects the specifications of the minimum input requirements needed to produce designated quantities of output, given the available technology. The traditional production theory emphasizes inputs that are directly related to outputs in the production process with certain technology, in which all of these factors are controllable by management.
This theory disregards the external factors beyond the control of firms that influence the relationship behavior between input and output, and the technological progress. This paper aims to examine the impact of externalities on the performance of firms in the wood furniture industry in Central Java. For this purpose, production functions are estimated to analyze the effect not only of the agglomeration economies but also the network with international partners.
The analysis in this paper is conducted using annual manufacturing surveys from 1994 to 2003 collected by the Indonesian Central Statistical Bureau.2 Section 2 describes the research hypotheses and Section 3 explains the research methodology. The profile of wood furniture firms in Central Java is described in Section 4. Results are presented in Section 5, followed by conclusions in Section 6.
2. Literature review and research hypotheses
2.1 Specialization The clustering of specialized firms offers benefits for firms inside the cluster that cannot be enjoyed by other firms. This notion goes back to the seminal work of Alfred Marshall and over time different mechanisms have been proposed and empirically tested. It is beyond the scope of this section to discuss this literature at length. All proposed mechanisms rely on the notion that clustering provides externalities and opportunities for joint action that can lead to increasing returns. Knowledge spillovers, opportunities for joint action and sharing a specialized labor force feature prominently in the list of arguments.
Conceptually, these were neatly summarized by Duranton and Puga (2004) referring to the benefits of learning, sharing and matching. Based on these insights, we hypothesize that the clustering of the specialized firms has a positive effect on firm performance. More specifically, we will investigate whether the clustering of L&M specialized firms has a positive effect on firm performance and whether the clustering of small scale specialized firms has a positive effect on firm performance.
2.2 Diversity In a strongly related line of research, several studies have documented that urban location provides benefits of the clustering of diverse firms. Research on the impact of urban location confirms that a city location increases firm productivity (Venables, 2005). Most research on the advantages of urban cities emphasizes the contribution of the advantages of labor (Glaeser and Maré, 2001). However, cities also create an environment conducive to innovation through knowledge spillovers (Acs, 2002). According to Webster and Muller (2000), the competitiveness offered by cities is derived from its unique economic structure, human resources, the institutional environment, and territorial endowments. Rosenthal and Strange (2003) argue that, despite labor market pooling, input sharing, and knowledge spillovers, cities also provide natural advantages, home market effects, consumption opportunities, and rent-seeking advantages. However, in many developing countries, cities have developed as centers of poverty and social collapse (Webster and Muller, 2000), as crime and violence increasingly affect the lives of the BPS distinguishes firms into large (L), medium (M) and small (S) firms. A large firm is a firm that employs at least 100 workers; medium firms employ between 20 and 99 workers; and small firms employ 5 to 19 workers.
people. Consequently, there is a tendency among some scholars to explore the potential of developing small cities (Rondinelli, 1983) giving rise to the fundamental question of optimal city size. In view of these arguments, our second hypothesis therefore is that the clustering of diverse firms has a positive effect on firm performance.
2.3 Exports There is a plethora of research showing that exporting producers have higher productivity than nonexporters (e.g., Aw and Hwang, 1995; Bernard and Jensen, 1995; Aw et al., 1997; Clerides et al., 1998;
and Aw et al., 2000). The self-selection theory and learning-by-exporting provide important explanations for this better performance (e.g., Clerides et al., 1998; Bernard and Jensen, 1999). A study on Spanish manufacturing firms (Delgado, 2002) confirms the hypothesis that productivity for exporting firms is higher than for non-exporting firms. However, they find evidence supporting the self-selection argument, whereas learning-by-exporting is rather weak. Meanwhile, research on two new industrialized countries (Aw et al., 2000) provides different evidence. In Taiwan (China), self-selection models can predict variations in productivity, whereas in Korea no statistically significant difference is found in productivity between firms that enter or exit from the export market. As the majority of research indicates that firm exporting has a positive effect on firm performance, further research to test this thesis is needed. Our third hypothesis is therefore that an exporting firm performs better than a non-exporting firm.
2.4 Foreign ownership The hypothesis that foreign-owned firms perform better than domestic firms is supported by much research (e.g., Asheghian, 1982; Kumar, 1984; Grant, 1987). Studies in the US (e.g., Doms and Jensen, 1998; Howensteine and Zeile, 1994) provide evidence that the source of better performance in foreignowned firms is their significantly higher labor productivity than those remaining under domestic ownership, since these firms spend more in employee investments. Howensteine and Zeile (1994) argue that foreign-owned establishments are more capital-intensive and larger. Meanwhile, research in the UK (Griffith and Simpson, 2003; Criscuolo and Martin, 2003) gives the same results. According to Bellak (2004), the difference in performance between foreign firms and domestic firms is caused by differences in productivity, technology, profitability, wages, skills, and growth. Moreover, Douma et al. (2006) argue that the reasons for better performance are due to larger shareholding, higher commitment, and longerterm involvement. The same results are also evident in research on developing countries. Studies in Indonesia (Arnold and Javorcik, 2005) support the above findings. Studies in developed countries provide evidence that better performance is due to the multi-nationality of the firm, rather than the nationality of the firm owner, since foreign ownership is much less important. We hypothesize that foreign-owned firms perform better than non-foreign-owned firms.
3. Research methodology
3.1 Model specification We now turn to a discussion of the model specification used for the estimation. A production function is used to describe the transformation process in accordance with which inputs are transformed into output, taking into account the contributions of external factors to productivity (cf. Moomaw, 1983; Nakamura, 1985). The analysis rests on a production function that relates the output of firms for a given sector in a region to a number of variables according to Y = A·f (L, K) where Y is output, A is technology (modeled in a Hicks-neutral fashion), L is labor, and K is capital. Typical examples of production functions often used are the so-called Cobb-Douglas and CES production function, the former being a restricted version of the latter. Examples of studies on agglomeration economies employing the Constant Elasticity Substitution (CES) function to examine the impact of externalities in the manufacturing industry are Alperovich (1980) and Calem and Carlino (1990).
In this study we employ the CES production function. There are different versions of the CES function used in the literature, of which the Arrow et al. (1961) model specification has become the standard specification (see Klump and Preissler, 2000). The functional form adopted by ACMS is as
where α is a distribution parameter, ρ is a substitution parameter, and µ captures the economies of scale. In this model the elasticity of substitution is estimated along with dummies for cluster factors and international linkages, characterizing the determinants of A (on which we elaborate below).
In the case of a production function that is characterized by constant returns to scale, µ equals 1.
The elasticity of substitution between K and L is equal to 1/(1+ρ). We typically assume that ρ –1, to avoid the substitution elasticity from being negative. If the substitution elasticity is zero, the two input factors can be interpreted as perfect complements (the production function is then of the Leontief type). In the case of a substitution elasticity equal to 1, the production function is of the Cobb-Douglas type (viz.