A Jiang Gang and a Weng: How Hong Kong and Taiwan Aged the Ingredients for China's Rapid Growth

For most of the twentieth century, industrialization was treated as something countries did largely within their own borders. Governments built roads and power plants, firms built factories, workers learned trades, and national economies became more productive. China’s rise after 1978 is usually told in some version of that language: reform, cheap labor, infrastructure, foreign investment, and an enormous domestic market.

All of those things mattered. But they leave an awkward question sitting in the middle of the story. China did not merely become capable of making more things. It became extraordinarily good, extraordinarily quickly, at making things for people elsewhere: American retailers, Japanese electronics companies, European brands, and eventually nearly every kind of foreign buyer with a specification sheet and a delivery date.

That kind of competence has to come from somewhere. It is built out of routines, relationships, judgment, accumulated mistakes, and thousands of forms of practical knowledge that do not appear automatically when wages fall or regulations change. By the time “Made in China” became an ordinary description rather than an exotic one, a great deal of that learning had already happened.

Some of it had happened outside China. China had plenty of workers in 1980. It had factories, engineers, technical institutes, and an industrial base large enough to manufacture everything from machine tools to military aircraft. It also had a government increasingly willing to experiment with markets, welcome foreign investment, and devote enormous resources to infrastructure. What it did not yet have was much experience manufacturing ordinary goods for demanding foreign customers.

A factory supplying an American shoe company or Japanese electronics manufacturer has to do more than make shoes or circuit boards. Someone has to know how to quote the job, translate a customer's specifications into a production process, design or acquire the tooling, qualify suppliers, control defects, schedule production, manage inventory, package the product correctly, clear customs, and get it onto a ship when promised. If the customer changes the specification halfway through the run, someone has to know what that change does to twenty suppliers farther down the chain. None of this is especially mysterious. It is also not something a government creates by declaring an export-processing zone.

China's reforms made this kind of production possible. They do not by themselves explain how China became good at it so quickly.

Part of the answer was sitting just offshore.

Two Ways of Leaving China

Hong Kong entered the mainland manufacturing story first. Its postwar population had been swollen by refugees from the mainland, including industrialists from Shanghai who arrived with capital, machinery, technical knowledge, and commercial relationships. By the 1970s Hong Kong had become a formidable export-manufacturing economy in its own right.[1] When China began reopening, Hong Kong manufacturers had an almost absurdly convenient low-cost hinterland on the other side of the border.

During the 1980s, production spread from Hong Kong into Shenzhen, Dongguan, and the rest of the Pearl River Delta. By around 1990, tens of thousands of factories in the region were performing outward processing or operating as joint ventures associated with Hong Kong firms, employing millions of mainland workers.[2] Management, finance, marketing, and buyer relationships could remain in Hong Kong while increasingly large portions of production took place a short drive away.

This was already a massive transfer of industrial capability. Hong Kong was doing far more than financing and coordinating mainland production; it was extending a mature export-manufacturing system across the border.

Taiwan arrived at the same opening through a different historical path. In 1949, at the end of the Chinese Civil War, the Nationalist retreat carried the Republic of China state and much of its military establishment to Taiwan, along with officials, businessmen, professionals, technicians, intellectuals, families, refugees, and conscripts drawn from many parts of China. Something on the order of a million mainlanders arrived in Taiwan around the end of the civil war.[3]

An eighteen-year-old soldier who reached Taiwan in 1950 was forty-eight in 1980. A thirty-year-old was sixty. The population that arrived with the retreat had three decades in which to change occupations, accumulate experience, build businesses, create families, and form entirely new relationships.

Veterans were civilianized into agriculture, construction, manufacturing, trades, and other occupations. Government programs placed tens of thousands into civilian employment.[4] Veterans' productive enterprises themselves evolved from agriculture and handicrafts toward processing, manufacturing, construction, and more technically demanding industries. Meanwhile people became supervisors and salesmen, opened businesses, accumulated capital, and entered networks that had not existed in 1950.

Taiwan had changed even more dramatically. The Taiwan of the mid-1980s was the product of another thirty-five years of development. American aid and access to American markets mattered. Japanese equipment, technical relationships, and production practices mattered. Taiwanese government industrial policy and technical institutions mattered. So did foreign OEM relationships, domestic entrepreneurship, returning engineers, and several decades of learning how to survive as an exporter.[5]

Separation was not merely the circumstance under which that development occurred. It helped determine its direction. Taiwan had been cut off from the enormous mainland hinterland into which an expanding manufacturer might otherwise have grown. Its domestic market was small. Firms that wanted to become much larger therefore had to sell beyond Taiwan, which meant submitting themselves to customers who had no particular reason to buy Taiwanese goods. American, Japanese, and European buyers could demand lower costs, better quality, reliable delivery, and increasingly sophisticated production, then take their business elsewhere if Taiwanese suppliers failed to provide them.

That was a severe filter. Firms that survived it accumulated precisely the capabilities that would later matter in China: production discipline, supplier coordination, quality control, responsiveness to engineering changes, export logistics, cost control, and familiarity with demanding foreign customers. Taiwan's separation from the mainland denied its manufacturers an easy continental market and made foreign competition unusually consequential to their development.[6]

By the mid-1980s, Taiwan had become a dense manufacturing economy populated by small and medium-sized firms that knew how to work with one another. It had machine shops, mold makers, component manufacturers, assemblers, trading companies, engineers, purchasing managers, salesmen, and factory supervisors. Its firms knew foreign buyers and had spent years learning what those buyers would and would not tolerate.

Finding a Way Back

The first Taiwanese manufacturers to enter mainland China were not Foxconn-sized companies betting enormous amounts of capital. They were smaller and harder to see.

Taiwan still restricted direct economic contact with the mainland, so investment could be routed through Hong Kong or other jurisdictions and disappear statistically into somebody else's foreign-investment column.[7] On the Chinese side, property rights, contracts, and administrative rules were still unreliable enough that a perfectly legal agreement was not necessarily a sufficient reason to put a factory somewhere.

Under those conditions, relationships carried information that institutions could not yet provide.

Fujian and Guangdong dominated the early geography of Taiwanese investment, precisely where family, native-place, dialect, and business connections were easiest to activate. In 1991, roughly three-quarters of recorded Taiwanese investment went to those two provinces. Research on Taiwanese location decisions finds that ethnic and native-place ties materially affected where firms chose to invest.[8]

Family connections were one part of a larger relational infrastructure that included native-place ties, dialect communities, business introductions, intermediaries, and relationships with local officials. These connections reduced uncertainty before formal institutions could do the job reliably.

The early firms were therefore important out of proportion to their capital. They discovered whether a Taiwanese-run factory could function on the mainland. They learned which officials could be trusted, how land arrangements actually worked, whether electricity stayed on, how workers were recruited, how customs behaved, what happened during a dispute, and whether money could be made and extracted. Their existence also changed the question facing the next Taiwanese businessman. He could ask someone who was already doing it.

By the end of the 1980s, the economics on both sides of the Strait were changing quickly. China was offering cheap labor, inexpensive land, special economic zones, increasingly aggressive local governments, and greater policy tolerance for foreign manufacturers. Taiwan was experiencing rising wages and land costs, labor shortages, currency appreciation, environmental pressure, and increasing political tolerance of economic contact with the mainland.[9]

The pressure gradient had changed at the same time that the pioneers were reducing uncertainty about what lay on the other side. Shoes, garments, toys, plastics, bicycles, umbrellas, and other labor-intensive industries began moving at much larger scale. These companies still relied heavily on relationships, but relationships no longer had to carry the entire institutional burden. Policy had begun to carry some of it too.

A factory moved, and over time its neighborhood formed around it.

Three Hundred Suppliers

Delta Electronics established operations in Shijie, Dongguan, in 1992. In the years that followed, a cluster of roughly three hundred Taiwanese suppliers with which it had already worked in Taiwan formed around the operation. The larger Dongguan supply network eventually numbered well over a thousand firms. In interviews with Taiwanese manufacturers, more than half identified the movement of an upstream supplier or downstream customer as a principal reason for their own investment in China.[10]

An assembler already knows which mold maker can hold a tolerance, which component supplier can increase volume without destroying quality, which company can turn an engineering change quickly, and which supplier's promise of delivery on Thursday actually means Thursday. The suppliers know the assembler's engineers and purchasing people. They understand its documentation and expectations. They may already be supplying other firms in the same network.

Move enough of those relationships together and a factory has not merely changed addresses. A production system has moved.

Researchers comparing Hong Kong and Taiwanese electronics firms operating in the same Dongguan environment found materially different organizational patterns. The Hong Kong firms tended toward smaller, more owner-centered organizations and greater reliance on short-term market contracting. Taiwanese electronics firms developed denser interfirm production networks. The same Chinese labor market, local government, infrastructure, and national policy regime were hosting different industrial organizations imported from outside.[11]

When Taiwanese manufacturers reported that their mainland supply chains were becoming “localized,” many of the sophisticated local suppliers were themselves Taiwanese companies that had moved to China. Indigenous Chinese firms were often still supplying packaging and other relatively low-value inputs.[11]

China's first encounter with these networks was therefore transplantation rather than immediate absorption. A functioning Taiwanese production system had been transplanted onto Chinese territory.

China's exports could explode while the Chinese value added in some processing exports remained relatively modest. Components, machinery, engineering knowledge, supplier relationships, and sometimes much of the management system still came from outside China. The Chinese contribution might initially consist disproportionately of land, labor, infrastructure, and increasingly competent factory operations.

Transplantation came before absorption. The transplanted system then began to change in China. Chinese workers became supervisors. Technicians became engineers and managers. Local firms learned what foreign assemblers required and qualified for progressively more difficult work. Employees moved between firms or started their own. Tooling and component capabilities that initially had to be imported could increasingly be supplied locally. The boundary between a Taiwanese production network operating in China and a Chinese industrial ecosystem consequently became less distinct with time.

The first Taiwanese investor arrived in Kunshan around 1990. Early investors developed relationships with local officials and then promoted the location to other Taiwanese businesses. WUS Printed Circuit became an important early anchor; its founder is credited in accounts of Kunshan's development with helping recruit additional Taiwanese companies. Suppliers followed customers, customers pulled suppliers, business associations formed, and local officials learned what Taiwanese manufacturers required.[12]

The Chinese term guanxi refers to the network of personal relationships through which trust, introductions, obligations, information, and favors can move. It is sometimes translated simply as “connections,” but that misses the reciprocal character of the relationships: guanxi is not merely knowing someone useful, but being embedded in a network in which past and future dealings give the relationship weight.

Kunshan marked a different stage in the process. Unlike the early movement into Fujian and Guangdong, Taiwanese firms there were no longer relying primarily on ancestral or native-place ties to establish a foothold. The network had become capable of creating the relationships it needed as it moved.

The first Taiwanese entrants into China had benefited from inherited guanxi. By the time the network reached places such as Kunshan, it could manufacture guanxi of its own. A successful factory created relationships with officials, workers, suppliers, customers, landlords, bankers, and other Taiwanese manufacturers. Those relationships reduced the risk for the next firm just as an uncle or native-place connection might have reduced it for an earlier pioneer.

The channels were no longer limited to the ones history had left behind. The industrial system could create new ones as it moved.

How Large Was This?

During the 1990s, China's merchandise exports increased from roughly $63 billion to $249 billion. Foreign-invested enterprises went from producing about one-eighth of Chinese exports at the beginning of the decade to nearly half by its end. Their exports alone increased by roughly $110 billion, accounting directly for around sixty percent of the decade's increase in merchandise exports. Contemporary scholarship went further, attributing most, and by some estimates nearly all, of the growth of China's exports during the 1990s directly or indirectly to foreign-invested enterprises.[13]

This foreign-invested manufacturing economy was initially dominated neither by American multinationals nor by European ones. Hong Kong and Taiwan together accounted for very large majorities of China's foreign direct investment during important parts of the late 1980s and early 1990s—close to four-fifths in some years. Their share declined as China became a more conventional destination for investment from the rest of the world, but the early concentration is difficult to dismiss.[14]

The official Taiwan number understates an unknown amount of the Taiwanese contribution because Taiwanese firms frequently routed investments through Hong Kong while direct mainland investment remained politically restricted.[15] Published nationality columns therefore cannot tell us precisely how much Taiwanese capital was involved.

But Capital is only part of what we are trying to measure. A $500 million factory is easy to see in an investment table. A company that moves and causes thirty specialized suppliers to move with it may be economically more consequential than its initial capitalization suggests. The accumulated knowledge embodied in a production engineer has no FDI value. Neither does an established relationship with an American buyer, or knowing which supplier can make a difficult component reliably, or the ability of one Taiwanese businessman to persuade five others that Kunshan is a safe place to build.

By the end of the 1990s, foreign-invested firms in China were importing far more from Taiwan than they were exporting back to it. Taiwan was supplying components and intermediate goods into factories whose finished production was headed elsewhere. Hong Kong simultaneously remained a major intermediary through which Chinese exports reached developed markets.[16]

The China Circle was real. Hong Kong and Taiwan interacted with one another and with the mainland, and Hong Kong sometimes served as part of the plumbing through which Taiwanese capital entered China.

The two islands also represented different industrial networks. Hong Kong pioneered the enormous extension of an export-manufacturing economy into its adjacent Guangdong hinterland. Taiwan spent the decades of separation developing unusually dense supplier systems and production capabilities that proved capable of moving in clusters. As electronics became increasingly important during the 1990s, that difference mattered more.

China's broader economic growth still had many other engines. Agricultural reform produced enormous gains before Taiwanese investment became substantial. Township and village enterprises drove a large domestic industrial expansion. China had inherited a significant, if distorted, Mao-era industrial base. Infrastructure investment, urbanization, education, SOE reform, domestic entrepreneurship, and extraordinary rates of capital formation all contributed to China's growth.[17]

The narrower claim is substantial enough. China's emergence as a global export-manufacturing center depended heavily on production systems that had developed outside mainland China, and Hong Kong and Taiwan supplied an unusually large portion of the capital, organization, relationships, and industrial networks through which the early transformation occurred.

In the initial drafts of this essay, I wrote as if Taiwan and Hong Kong had acted as reservoirs of skill and capital. But that ignores critical changes that occurred in the two island states between 1949 and China's reopening. A reservoir stores something and later releases it. The large ceramic vessels used throughout China suggest a better image.

A jiàng gāng (酱缸) is a vat used to age and ferment sauces; a wèng (瓮) is a related earthen vessel used for storage and fermentation.[18] The material put into either vessel already has value. The interesting part is what happens while it is separated from the surrounding environment.

Hong Kong and Taiwan were not identical vessels. They contained different mixtures of people, institutions, capital, knowledge, and relationships, and they spent the decades of separation exposed to different markets, technologies, institutions, and constraints. What had been separated from China was therefore not merely preserved offshore. Old capabilities changed, old relationships survived or were reconstituted, and new networks formed around factories, suppliers, engineers, foreign buyers, and managers.

By the time mainland China became accessible again, those offshore systems contained capabilities that had not existed when separation began.

The scale of the transfer is difficult to recover from ordinary economic statistics. GDP statistics can measure value added. Trade statistics can measure a container crossing a border. Foreign-investment statistics can try to identify who supplied the capital, although routing through Hong Kong already makes even that less straightforward. None has an obvious column for a supplier following its customer, thirty years of accumulated factory judgment, a relationship with a foreign buyer, or the fact that three hundred companies already know how to work together.

Deng's reforms changed the conditions under which those capabilities could reconnect with the mainland. Managers and engineers arrived with experience accumulated elsewhere. Suppliers arrived with relationships already tested elsewhere. Factories brought production routines, tooling knowledge, buyer expectations, and habits of coordination that had taken decades to acquire. Chinese workers and firms then began the slower process of absorbing and extending those capabilities for themselves.

Much of China's growth still has to be explained elsewhere. But the extraordinary speed with which its coastal export economy became capable of supplying the world is less mysterious once we notice how much of the learning had already taken place offshore. When the channels reopened, some of that accumulated capability crossed into China in factories and investment accounts. Much of the rest came less conspicuously, in the people, relationships, toolboxes, and sample cases that made the factories work.

Notes

[1] Hong Kong Memory, “Industrialization in Postwar Hong Kong: The China Factor”. The account notes that more than thirty cotton-spinning industrialists from Jiangsu and Zhejiang came to Hong Kong between 1945 and 1950, bringing machinery, capital, and modern management practices; it also describes the broader contribution of postwar mainland migration to Hong Kong manufacturing.

[2] International Monetary Fund, “Growth, Structural Change, and Economic Integration”. The IMF cites an estimate that by the end of 1991 Hong Kong firms operated about 14,000 foreign-funded enterprises employing roughly three million workers in Guangdong alone. A contemporary INSEAD study similarly estimated 20,000 outward-processing factories plus 7,000 Sino-Hong Kong joint ventures in the Pearl River Delta in 1990.

[3] Dominic Meng-Hsuan Yang, “Together in the Same Boat: Exiled Nationalist State and Chinese Civil War Exiles in 1950s Taiwan”. Yang describes roughly one million mainland Chinese displaced to Taiwan with the Nationalist regime and emphasizes that the population included common soldiers, petty civil servants, and refugees from varied walks of life rather than only political and economic elites.

[4] Yang, “Together in the Same Boat”. By 1966, Taiwan's Veterans Affairs Council had helped 71,311 veterans find new employment and released roughly another 50,000 into the general population; veterans were also placed in farming and public-infrastructure projects.

[5] Ken C. Y. Lin, International Monetary Fund, “Industrial Development and Changes in the Structure of Foreign Trade: The Experience of the Republic of China in Taiwan, 1946–66”; and World Bank, “Small-Scale Industries in Two Rapidly Growing Less Developed Countries: Korea and Taiwan”. Both describe Taiwan's rapid shift from an aid-dependent, agriculture-heavy economy toward export-oriented manufacturing; the World Bank notes that by 1976 manufacturing produced more than 40 percent of GDP and more than 90 percent of exports.

[6] Lin, “Industrial Development and Changes in the Structure of Foreign Trade”. Lin explicitly identifies the small size of Taiwan's domestic market as a constraint on continued import-substitution growth and describes the resulting turn toward manufactured exports and developed-country markets. The “filter” language in the essay is an inference from this economic structure, not Lin's terminology.

[7] The China Quarterly, “Economic Integration within Greater China: Trade and Investment Flows Between China, Hong Kong and Taiwan”. The study notes the acceleration of Taiwanese investment after Taipei permitted visits to mainland relatives in 1987 and the subsequent 1991 approval of mainland investment conducted through a third country. Contemporary reporting also documents Taiwanese investments routed through Hong Kong and other jurisdictions.

[8] Peterson Institute for International Economics, “China-Taiwan Economic Relations”, attributes the early dominance of Guangdong and Fujian in Taiwanese investment partly to family ties as well as export-processing policies and the fit with Taiwan's SME manufacturing base. A 1991 Taiwanese government survey reported Guangdong and Fujian accounting for 88 percent of projects and 67 percent of investment value; another contemporary compilation put their cumulative shares at 50.1 and 23.9 percent respectively. The exact percentage depends on whether projects, approved amounts, cumulative investment, or locally reported investment is being counted.

[9] U.S. Department of State, “Taiwan: Investment Climate”, describes Taiwanese firms beginning to relocate production to China in the late 1980s to exploit lower labor and land costs. Contemporary reporting likewise records Taiwanese manufacturers moving production as wage differentials widened. The essay's “pressure gradient” combines these Taiwan-side pressures with China's simultaneous liberalization and local investment incentives.

[10] Chun Yang, “Divergent Hybrid Capitalisms in China: Hong Kong and Taiwanese Electronics Clusters in Dongguan”, and Yang, “Backward Linkages of Cross-Border Production Networks of Taiwanese PC Investment in the Pearl River Delta, China”. Delta invested in Shijie in 1992; the later cluster included approximately 300 Taiwanese suppliers. The “three hundred suppliers followed us” formulation comes from a 2006 interview and should be understood as describing the formation of the cluster rather than a single 1992 relocation event. Delta's own 50th-anniversary history independently dates its first mainland factory to Dongguan in 1992.

[11] Yang, “Divergent Hybrid Capitalisms in China”. The comparative study examines Hong Kong and Taiwanese electronics firms in the same Dongguan environment and documents differences in firm organization, interfirm networks, and sourcing. It also distinguishes nominal localization from sourcing through Taiwanese suppliers that had themselves relocated to China.

[12] Kang-tsung Chang, Konstantinos A. Melachroinos, and Ching-mu Chen, “Beyond Technological Relatedness: An Evolutionary Pro-Growth Coalition and Industrial Transformation in Kunshan, China”, describes Kunshan's first Taiwanese investor in 1990, WUS Printed Circuit's major 1992 investment, and founder Wu Ligan's role in promoting Kunshan to other Taiwanese firms. Research on global production networks and local institution building in Suzhou likewise describes WUS as an anchor and Wu's close involvement with Kunshan officials.

[13] International Monetary Fund, “China's Trade Flows”, records foreign-funded enterprises' share of Chinese exports rising from 12.6 percent in 1990 to 40.5 percent in 1997. An IMF review of foreign direct investment in China reports that the FIE share reached 45 percent by 1999 and that FIEs accounted for about half of total export growth from 1985 to 1999. Edward Graham and Erika Wada's survey of the literature, “Foreign Direct Investment in China: Effects on Growth and Economic Performance”, notes estimates attributing almost all Chinese export growth since 1992 directly or indirectly to foreign-invested enterprises.

[14] K. Honglin Zhang, “Why Does So Much FDI from Hong Kong and Taiwan Go to Mainland China?”, China Economic Review 16 (2005), 293–307. Zhang reports combined Hong Kong and Taiwanese shares of utilized Chinese FDI of 79.5 percent in 1987, 79.9 percent in 1988, and 79.6 percent in 1992, followed by a gradual decline as investment from other sources expanded.

[15] Graham and Wada, “Foreign Direct Investment in China: Effects on Growth and Economic Performance”, discusses the difficulty of identifying the ultimate source of investment recorded as originating in Hong Kong. Contemporary reporting in the South China Morning Post, “Anxious Taiwanese Investors Put Faith in Cross-Strait Political Crisis Blowing”, likewise notes Taiwanese investment routed through Hong Kong, Singapore, the Virgin Islands, and other jurisdictions.

[16] Zhang, “Why Does So Much FDI from Hong Kong and Taiwan Go to Mainland China?”, documents the regional trade structure associated with foreign-invested enterprises. A later U.S. State Department retrospective, “Taiwan: Investment Climate”, describes China-based Taiwanese factories processing Taiwan-made inputs into finished goods for export to the United States, Japan, Europe, and other markets. On Hong Kong's intermediary role, see The China Quarterly, “Economic Integration within Greater China”.

[17] International Monetary Fund, “Why Is China Growing So Fast?”, summarizes the contribution of agricultural reform, township and village enterprises, greater enterprise autonomy, private ownership, foreign investment, and export expansion to post-1978 growth. The essay's argument concerns one important part of this broader transformation rather than an exclusive explanation of Chinese GDP growth.

[18] The Chinese term 酱缸 (jiàng gāng) denotes a large pottery container used for making or storing soy sauce and other fermented sauces; see Wiktionary's 醬缸 entry. Taiwan's Ministry of Education defines 瓮 (wèng) as an earthen vessel used to hold things; see the Revised Mandarin Chinese Dictionary entry for 瓮. The fermentation comparison in the essay is metaphorical; the two vessel terms are not being assigned one-to-one to Hong Kong and Taiwan.