Will Shanghai be the Detroit of China? Michael Smitka Professor of Economics Washington and Lee University Lexington, VA USA

Transcription

1 Will Shanghai be the Detroit of China? Michael Smitka Professor of Economics Washington and Lee University Lexington, VA 50 USA Early and Incomplete Draft For Presentation at the Industry Studies Association Seattle, Washington May 0 - June, 08 Abstract China is the largest automotive market in terms of both production and sales, turning out million cars and trucks in 0, while NAFTA accounted for only. million and Europe 8. million. As the market matures, will production coalesce around an auto ally or auto corridor, as happened in both Europe and North America over the past 0 years? Or will the size of the market allow multiple centers to coexist? This paper does an initial analysis looking at automotive suppliers, and concludes that greater Shanghai region is by far the largest region, accounting for about half of all parts firms, while both the southern area of Guangdong Province and the region of Beijing and Manchuria each account for / th. That may be enough to permit economies of scale on a regional basis. Draft 5//8, : AM page of

2 Introduction The manufacturing of automobiles is a weight-gaining process. As emphasized by James Rubenstein this has two implications. One is that final assembly will occur close to the customer, captured in the geographic center of gravity of sales. The second is that component production will cluster around assembly. He and Thomas Klier demonstrate observed both of these forces at work in North America and in the European Union. (Rubenstein ; Klier and Rubenstein 008, 05). Other forces are at work. In Europe and NAFTA, the periphery (Mexico, Eastern Europe) drive smaller vehicles, and so the locus of assembly of such vehicles has shifted closer to those markets. Similarly, the rise of global platforms and lower costs of trade have led to the globalization of the supply chain; the production of some parts has moved further away, not closer (Klier and Rubenstein 00). Tariffs also matter, leading Audi to locate its sole production in the Western Hemisphere in Mexico, where its market is very small (Smitka and Warrian 0, Chapter ). How will the geography of assembly and parts production evolve in China? Until the Chinese market resembled that of Europe prior to Common Market. The economy was highly decentralized, with strong provincial and Tier I city support of local industries. As a result, at least of China s provinces had their own truck producers, and there were also widely scattered passenger car factories. Indeed, at peak there were some 00 assemblers in what was still a million unit market (Smitka 0, Chapter 5). In Europe the industry realigned as trade barriers fell, while in the US branch plants on the East and West coasts all closed, giving rise to a central auto ally or auto corridor in which most assembly and parts production concentrated. Will we see the same process occur in China? In particular, will Shanghai become a new Detroit or Stuttgart? This paper represents an initial attempt to capture the geographic locus of suppliers, drawing upon an annual directory of the Chinese industry. I do not address the geography of final sales, nor have I seen any geographic analysis in my literature survey. Draft 5//8, : AM page of

3 The Chinese Automotive Industry China adopted an import substitution industrialization strategy in the 50s, with the initial push through a set of 5 plant projects built with Soviet technical and financial assistance. For the automotive industry, this took the form of the single plant of the First Automotive Works in Changchun in Jilin Province in Manchuria. Production of trucks commenced in 5, and of cars in 58. This was followed by a mirror plant outside Wuhan in Hubei Province (the forerunner of today s Dongfeng Motor), and truck assembly at a military arsenal in Chongqing (which evolved into Chang an Automobile). Over time additional locations added low-volume operations. By the end of the 0s there 0 such plants scattered across China, and 0 by 80, all with captive markets in firms owned by provincial and city governments. Change began in the mid-80s, as economic growth accelerated the demand for motor vehicles, both trucks and cars. This led to new entry and (in the face of scarcity) legal and grey-market imports, which in 85 exceeded domestic production. In response the government permitted a few joint ventures to kickstart the industry and to save on foreign exchange. VW and FAW, for example, set up a plant in Changchun to make the Audi 00 using an antiquated production line imported from South Africa. Other ventures were set up in Guangzhou and Beijing; though neither did well. The most successful venture began production in 85 as a joint venture between the Shanghai Automotive Industry Corporation (SAIC) and VW to produce the Santana, starting with the assembly of CKD kits with an increase in local content over time. Shanghai was one of the most prosperous cities in China, and the local taxi company provided a captive market; local protectionism meant that no Santanas were to be found in Guangzhou, Chongqing, Beijing, Hubei or Jilin. The challenge of developing local content and the lack of competition meant that in 000 Shanghai VW continued to produce a model first launched in Germany in 8. Only when SAIC set up a joint venture with GM in 8 did VW begin work to produce a more modern vehicle. (Thun 00; Gallagher 00; Smitka 0). Draft 5//8, : AM page of

4 At the start of the 000s there was thus modest production in Changchun, Jilin; Beijing; Guangzhou, Guangdong; Chongqing, and Wuhan, Hubei, and a much larger nexus in Shanghai. The long history of heavy industry in Manchuria meant that there were some suppliers there, but the low level of production and lack of government support meant that there was only a very small supply base in Guangdong, Chongqing, Hubei and Beijing. There were also a small number of parts ventures aimed solely at export were set up in scattered locations by global suppliers, but these had little or no link to local car companies (author interviews with assorted global suppliers). Only the Shanghai area had a nexus of components manufacturers geared to the demands of the up-to-date models that the various joint ventures launched after GM commenced production of the Buick Sail in 00 (Thun 00). In the now-standard Chinese analysis, subsequent development resulted in the formation of automotive production regions, each centered on multiple joint ventures, such as those of Chang an with Suzuki, Ford and Mazda in central China, though as the summary will note, the geography of assembly plants is more complicated than that. In any case, this leads to the core hypothesis of the paper, that each of the regions has its own localized group of suppliers. Literature on Chinese suppliers and on the geography of the industry in China [to be added. the literature is very thin. even though suppliers likely employ out of workers in the automotive industry, there is very little specific to suppliers in English, Japanese or Chinese, much less to their geography. overview of Japanese, English and Chinese papers] Analytic Approach Data As a data source I use of the 0 edition of the annual industry directory assembled under the auspices of CAAM, the China Association of Automotive Manufacturers. This provides an entry for each supplier, along with cursory description of what each produces and often an indication of who their customers are. Draft 5//8, : AM page of

5 In addition, the directory provides a list of suppliers of each of a wide array of components. I analyze 5 engine component categories, 8 transmission and chassis categories, and 0 electronics categories, for a total of different types of components. I have not included capital equipment ( categories) and materials ( categories) as potentially less tied to automotive production locations, and have not added body-related suppliers ( categories) or interior suppliers ( categories). As an industry directory, it likely captures firms that focus only on the aftermarket or on specialized work vehicles (dump trucks, bulldozers). However, spot checks located internet descriptions for a couple dozen firms, and the additional information found thereby suggests that most of the firms in the directly remain in business and are indeed broadly automotive in focus. A bigger problem is that the directory does not segregate production for assembly from aftermarket production. That likely leads to an over count for items such as brake pads and calipers, but is less likely to be a problem for electronics. A larger problem is that the focus is on the headquarters location. Because Chinese suppliers for administrative reasons may set up subsidiaries in each province in which they operate, a few will be captured. For example, just-in-time production (and the vagaries of assembly scheduling) would seem to mandate that there be one or more seating suppliers near to each assembly plant. However, a check of the seating supplier category fails to turn up multiple plants for Lear and Adient or Japanese-named seating suppliers. Likewise, Hella s (Chinese-language) web site notes production locations in China in different provinces. So far I have only located the parent company entry in the directory. There is thus also an undercount. In the case of seat assembly there are plants are typically tied to particular customers, and so they do not directly contribute to a larger regional supplier base, but that will not be the case for every global or large Chinese domestic suppliers. In sum, my counts do not distinguish between production and aftermarket suppliers, or firms that only export. I do not distinguish between suppliers for passenger cars versus commercial trucks versus specialty vehicles. I in general fail to count factories for multi-establishment enterprises. However, it does Draft 5//8, : AM page 5 of

6 appear that the firms in the directory remain in business and with an automotive focus. Finally, suppliers are often multi-product. The same supplier thus will often be listed in more than one category. Not accounting for duplicates, I tabulated data on,85 suppliers of chassis and transmission components,, suppliers of engine components and, suppliers of electrical components. Geographic Classifications The received analysis of automotive production divides the country into centers, on the basis of groups of assemblers. However, that fails to capture whether the location of a supplier might be such that it could reasonably transport output to more than one center. I thus used Google Maps to calculate the time it takes to go by expressway from cities in which there were large numbers of suppliers to one or another of these centers. I used 5 hours as a cutoff for close, allowing a round-trip by a driver/truck, and hours for near enough to reliably schedule daily deliveries with a single truck and driver. These are arbitrary, and they are from city center to city center rather from the more likely outlying location of actual plants. It nevertheless provides an important check on centers. I used cities in these regions, completely ignoring the many truly remote locations with automotive production, such as Urumqi or the island of Hainan. (See the matrix in the data appendix.) With this process I generated the following groupings. One province with a substantial number of suppliers, Shandong, can potentially supply both the greater-beijing region and the Shanghai region. The central region of Chongqing and Sichuan, the south region of Guangdong, are truly apart. There is some overlap between Manchuria and the Beijing-Tianjin region. A large number of provinces feed into the Yangtze River basin. Northern Fujian (Fuzhou) is barely at the southern edge of the Yangtze basin, while the southern section (Xiamen) is within hours of Guangzhou in Guangdong. However, for many components there were no manufacturers in Fujian, and only in two categories did suppliers there account for a double-digit percentage of the total. I thus exclude it from the analysis. Nine other provinces with suppliers that I Draft 5//8, : AM page of

7 exclude because they lack geographic proximity to any of the centers are Shanxi, Inner Mongolia, Henan, Hainan, Yunnan, Guizhou, Shaanxi, Ningxia and Xinjiang. No suppliers were listed in the directory in the provinces of Tibet, Qinghai, and Gansu. The analysis thus focuses on 8 of China s provinces. Analysis The clear result is that the lower Yangtze basin accounts for the majority of suppliers in most of the categories I tabulated. The western region appears marginal, nor more in many categories than the total for the 0 provinces whose counts I excluded. The north (lumping Manchuria and the Beijing-Tianjin- Hebei nexus) each account for about / th of the total. The north is more prominent for purely mechanical components, which might be in line with the types of parts used before the explosion of joint ventures and modern vehicles that began in 00. There are clearly not regional centers for parts production, unlike the normal groupings used in discussing automotive assembly. Shanghai does appear to be the Detroit of China. Several anomalies and unusual patterns stand out. One is the large share of Zhejiang Province, which alone accounted for 5% of the entire set of engine component categories, % of the suppliers in the engine and transmission categories, and 0% for electrical component categories. I have not sampled a sufficient number to be sure, but this large share may reflect the Wenzhou industrial district model of large numbers of firms grouped together, making a similar product. As such, they are more likely to have an orientation towards the aftermarket or to low-volume production for trucks and specialty vehicles. I thus also ran a tabulation using only half the number of Zhejiang firms in each category. It did not change the results. Second, the low number of suppliers in Shanghai is surprising. Both Shanghai VW and Shanghai GM are located in the (provincial-level) city, and as Thun (00) and Gallagher (00) note, the city provided funding for localization of parts production for their suppliers. I thus expected to see Shanghai as a major center; however, among the various categories, Shanghai accounted for 0% or more for only engine Draft 5//8, : AM page of

8 categories and chassis categories. Its role was much larger for electrical components, where it had 0% or more of firms in 0 of 0 categories and the top share for. Third, Beijing stood out for software-related categories (navigation systems) while Guangdong stood out for consumer-electronics related ones (55% of audio devices, % of GPS firms). For Beijing the presence in high-tech components reflects government industrial policy. For consumer electronics it reflects the legacy of Guangdong as an early center of export production (most recently of cell phones) that gave it a broad skill-set in this area. Fourth, at the level of the individual category a province will occasionally stand out with an unusually large number of firms. Shandong has % of flywheel makers, % of steel wheel firms, 5% of brake rotor manufacturers and % of tire producers. Anhui generally had - firms per category, but had firms making drive shafts and 0 making front axles and transfer cases. Chongqing had firms making steering gears. There is thus some agglomeration of production that in the Chinese context likely reflects spinoffs from an original successful producer. Indeed, within the firms from Zhejiang, there were sometimes no firms from Wenzhou, the archetype of production districts, while in other categories there were tens of firms from that one district. Conclusion An analysis of firm counts is less satisfying than one based on output measures, but this analysis does strongly suggest that the greater Shanghai area is becoming the Detroit of China. Further work drawing from the industry directory could help indicate how many firms have engineering centers there. Interviews would be helpful in this: for example, do engineers based in Shanghai take frequent business trips to outlying production centers? While I can read Chinese slowly, I do not have enough spoken language to carry out interviews. Second, does the location of suppliers reflect the geography of production? In China, the two largest groups of brands by sales are those of GM and VW, so it might make sense that parts production is Draft 5//8, : AM page 8 of

9 centered around their production base. On the basis of total vehicles assembled, do the north and the west each account for about / th of production? If so, then there is no clear agglomeration effect. Third, what of production? Do firms in Chongqing and Sichuan sell more of their output there than elsewhere? Given the strong pressures of localism in China s political economy, that would not be surprising, but since policy emanating from Beijing has worked to foster a national economy and deemphasize the local. Are sales of passenger vehicle brands becoming more national? Even so, Sichuan and Chongqing have a combined population of 0 million, while Guangdong has 08 million, all more than any single European country. Perhaps China is large enough to support more than one automotive center, and the Yangtze River delta will remain first among equals, but not dominate the market. Draft 5//8, : AM page of

10 Manchuria grouping Jilin Liaoning Heilongjiang Liaoning Jilin Liaoning Table : Geographic Groupings brown shading is the 5-hour cutoff, blue shading the -hour cutoff Heilongjiang Heilongjiang Jilin Tianjin Tianjin Beijing North China grouping Beijing Tianjin Hebei Tianjin Beijing Beijing Hebei Hebei Shandong Shandon g Shandon g Henan Henan Henan Liaoning Liaoning Liaoning Intermediate: Shandong Shandong Qingdao Beijing Tianjin Hebei Jiangsu Shanghai Zhejiang Anhui Hubei Yangtze Basin Grouping Shanghai Jiangsu Zhejiang Anhui Jiangsu Shanghai Shanghai Jiangsu Zhejiang Zhejiang Jiangsu Shanghai Anhui Anhui Anhui Zhejiang Hubei Hubei Hubei Hubei Shandong Shandong Fujian Shandong Hunan Hunan Hubei South China Grouping Guangdong Guangxi Hubei West-Central Grouping Changchun Chongqing Chongqing Changchun Draft 5//8, : AM page 0 of

11 North North plus Shandong Geographic Tabulations Table : Electrical % % Yangtze Yangtze plus Shandong Yangtze less 50% Zhejiang Yangtze plus Shandong less 50% Zhejiang South West Not in a regional grouping 58% % % % % % 5% North North plus Shandong Table : Chassis and Transmission % % Yangtze Yangtze plus Shandong Yangtze less 50% Zhejiang Yangtze plus Shandong less 50% Zhejiang South West Not in a regional grouping 8% 58% % % % 5% % North North plus Shandong Table : Engine % 5% Yangtze Yangtze plus Shandong Yangtze less 50% Zhejiang Yangtze plus Shandong less 50% Zhejiang South West Not in a grouping 8% 5% % 5% % % % Draft 5//8, : AM page of

12 Table 5: Geography Matrix of Travel Times Draft 5//8, : AM page of

13 Draft 5//8, : AM page of Table : Sample of Firm Count Tabulations Beijing Tianjin Hebei Shanxi Inner Mongolia Liaoning Jilin Heilongjiang Shanghai Jiangsu Zhejiang Anhui Fujian Jiangxi Shandong Henan Hubei Hunan Guangdong Guangxi Hainan Chongqing Sichuan Yunnan Guizhou Shaanxi Ningxia Xinjiang Total Total engine assemblies 8 cylinder, block. head, liner piston piston ring piston pins 8 air intake camshaft 5 5 crankshaft connectin g rod bushing

14 Bibliography Brincks, Corey, Boleslaw Domanski, Thomas Klier, and James M. Rubenstein. 08. Integrated Peripheral Markets in the Auto Industries of Europe and North America. International Journal of Automotive Technology and Management 8 (): 8. China Association of Automotive Manufacturers (), and Beijing Auto Technology News (). 0. (China Automotive Industry Enterprises & Administrative Organizations). 0 ed. China Communications Press Co., Ltd. ( ). Gallagher, Kelly Sims. 00. China Shifts Gears: Automakers, Oil, Pollution, and Development. Urban and Industrial Environments. Cambridge, Mass: MIT Press. Klier, Thomas H., and James M. Rubenstein Who Really Made Your Car? Restructuring and Geographic Change in the Auto Industry. Kalamazoo, MI: W. E. Upjohn Institute. Klier, Thomas H., and James M. Rubenstein. 05. Auto Production Footprints: Comparing Europe and North America. Federal Reserve Bank of Chicago Economic Perspectives (): 0. Rubenstein, James M.. The Changing US Auto Industry: A Geographical Analysis. London [England] ; New York: Routledge. Smitka, Michael J. 0. China s Auto Industry: Success despite Policy. International Journal of Automotive Technology and Management ():. Thun, Eric. 00. Changing Lanes in China: Foreign Direct Investment, Local Governments, and Auto Sector Development. New York: Cambridge University Press. Draft 5//8, : AM page of