Evolution of production networks and the localisation of firms: evidence from the Thai automotive industry

Abstract

In this article, we compile a new data set of firms in the Thai automotive industry and discuss the localisation of both automotive parts suppliers and assemblers in Thailand. We first review the factors affecting the development and location of the Thai automotive industry and examine how Thailand came up with the strategy to promote one-ton pickup trucks as the first ‘product champion’. We argue that the government played a significant role in creating a ‘sizable’ domestic market and building the essential infrastructure, resulting in the promotion of dynamic growth and a vibrant productive capacity. Then, applying Duraton and Overman’s continuous approach, this paper examines the localisation of automotive firms. Using the spatial distribution of manufacturing establishments as a benchmark, this study finds that all categories of parts suppliers were significantly localised. Moreover, it is revealed that statistically significant localisation takes place only within 150 kilometers with respect to the bilateral distances between suppliers.

Keywords:

• Automobiles
• localisation
• production network

JEL Classification:

• C12
• L62
• R12

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 Thailand produced 2.01 million vehicles in 2019, (retrieved from http://www.oica.net/category/production-statistics/2019-statistics/ on 20 December 2020).

2 Based on our interview with the Thai Automotive Institute representatives, the database has been compiled and updated, but the number of entries is less than those in the directory. In addition, the TAI database contains information about ownership structure, i.e. the share of ownership by nationality. This information is the same as the business registration information at the Ministry of Commerce.

3 In this study, we follow the auto parts classification at www.marklines.com, which classifies parts into 13 main, second and third categories. Each category consists of several parts/components and sub-components. We then compare and match the type of main products from the Thailand Automotive Industry Directory. However, we had to create additional categories of parts that are not in the list of the auto parts classification. They are automobile assembly, agricultural machinery and other transport machinery, chemicals, oil, lubricants, paint, etc., accessories, services (trading, logistics, trade shows, training, etc.), and machine tools, jigs & fixtures, moulds & dies, etc.

4 Geographical coordinates of postal codes are simultaneously used with those of districts. Therefore, the same geographical coordinates are used for both automobile and manufacturing firms.

5 Meanwhile, the local confidence interval is defined as follows: For each industry or category of industry, we rank our simulations in ascending order and select the 5th and 95^th percentile to obtain a lower 5% and an upper 5% confidence bound, for each kilometer in the interval. Following Duranton and Overman (2005), Nakajima, Saito, and Uesugi (2012), 180 km was used as the benchmark.

6 Manufacturing share in GDP of the Eastern Seaboard increased from 21–22 percent during 1981 and 1985 to 27–28 percent during 1990 and 1997.

7 The project received several sources of finance, including from the Japanese Official Development Assistance (ODA) and World Bank Loans (Watanabe 2003, p. 142). The Board of Investment granted higher zoning incentives to firms in these provinces and recently this area has been promoted under the Eastern Economic Corridor (EEC).

8 During the period of import substitution, both inputs and markets were provided by the metropolitan area, where suppliers and consumers reside. Thus, a metropolitan area, such as Bangkok, was the best location for import-substitution firms. However, once the market opens to international trade, the metropolitan area loses (at least a part of ) such advantages, whereas the frontier region becomes more attractive, especially for export-oriented firms, owing to superb access to imported inputs as well as to international markets (Krugman and Elizondo, 1996; Fujita, Krugman, and Venables 1999).

9 Most notably, Mitsubishi set up an assembly factory in the close vicinity of the Laem Chabang Port in Chonburi in 1992, followed by Ford–Mazda and General Motors, each of which established factories in Rayong in 1998 and 2000, respectively. Note that the assemblers that set up factories in Rayong and Laem Chabang were highly export-oriented, whereas Honda, which established a factory in Ayuthaya, had a lower dependency on exports (Kuroiwa, Bhandhubanyong, and Yamada 2016).

10 We also estimated the kernel densities of bilateral distances between firms, in the cumulative number, by period. Our calculation on the cumulative basis is consistent with the previous discussion, in which firms tended to locate in Bangkok and its suburbs prior to the mid-1980s. The spatial distribution of firms was characterized by a cluster of firms separated by an average distance of 44 kilometers (1st period) and 41 kilometers (2nd period), and a median of 29 kilometers and 28 kilometers. As time passed, the bilateral distances between firms became more dispersed. The average distance increased to 69 kilometers (3rd period) and 68 kilometers (4th period) and the median to 45 kilometers and 44 kilometers after the mid-1980s.

11 This is consistent with our results shown in Figure 3 that the highest peak of around 30 km remained in all periods.

Additional information

Funding

Financial support by the Economic Research Institute for ASEAN and East Asia (ERIA) for the compilation of the data on the automotive-related firms in Thailand is gratefully acknowledged.

Notes on contributors

Ikuo Kuroiwa

Ikuo Kuroiwa is Professor at the University of Niigata Prefecture, Japan. He holds a Ph.D. from the University of Pennsylvania. Before joining the university, he worked for the Institute of Developing Economies (IDE-JETRO). His research interests include spatial economics and input-output analysis. He edited books on economic geography in East Asia and published academic papers on agglomeration economies and manufacturing clusters.

Kriengkrai Techakanont

Kriengkrai Techakanont is Associate Professor of Economics, Faculty of Economics, Thammasat University. He received his doctorate from Hiroshima University, Japan in 2002. His research interests include automobile industry, technology transfer, global production network, cluster, and interactive learning. His current research is focused on circular economy and the end-of-life vehicles in Thailand.

Souknilanh Keola

Souknilanh Keola is an associate senior research fellow at Institute of Developing Economies, Japan External Trade Organization. He received bachelor degree in computer sciences and master degree in economics from Toyohashi Institute of Technology and Nagoya University respectively. He also held/holds adjunct lecturer positions at several universities, including Tokyo University of Foreign Studies, Chiba University, Keio University and currently Sophia University, Japan.