3 The making of industrial behemoths

Patterns of state intervention and industrial organization

 

Presciently anticipating theorists of the ‘developmental state,’ Friedrich List had argued in 1885 that in ‘less advanced nations … a perfectly developed manufacturing industry, an important mercantile marine, and foreign trade on a really large scale, can only be attained by means of the interposition of the power of the State’ (quoted in Leftwich, 1995: 401). Similarly, Alexander Gerschenkron (1962) concluded from a study of pat-terns of ‘late industrialization’ that state intervention in ‘relatively backward’ countries was especially important because the average size of plants needed to be larger precisely when capital was scare. Building on his analysis, Ellen Kay Trimberger's (1978: 4) comparative study of Egypt, Japan, Peru, and Turkey indicated that states were able to be effective in economic development when the

bureaucratic state apparatus achieved relative autonomy when, first, those holding high civil and military office were not drawn from dominant landed, commercial, or industrial classes; and, second, where they did not immediately form close relations with these classes after achieving power.

While a recognition of the significance of the relative autonomy of the state and the vital importance of government intervention in the economy for late-industrializers thus has a long and distinguished pedigree, the onset of the irreversible decline of centrally-planned economies since the late 1970s led to a neoclassical resurgence. It was argued that state intervention had ‘generated inefficient industries requiring permanent subsidization for their survival’ and that it also tended to foster ‘rent-seeking’ and thereby detracted ‘the attention of economic agents from productive activities into lobbying for increased allocations of government subsidies and protection’ (Onis, 1991: 109). The clearest empirical refutation against this revival of neoclassical orthodoxy emerged from the Asian rim-lands.

By the early 1980s, the sustained rates of economic growth registered by these states led to a reassessment of the role of dirigisme in economic development. Characterizing Japan as a ‘developmental state,’ Chalmers Johnson (1982: 18–20) distinguished it from both the ‘market rational’ orientation of other core economies and the ‘plan ideological’ orientation of centrally-planned economies. Unlike ‘market rational’ states like the United States which were concerned with the rules and procedures of economic competition rather than with the type of industries that were needed, the ‘plan rational’ orientation of the Japanese government emphasized substantive economic issues. Conversely, unlike the centrally-planned economies, that he charged were primarily concerned with the state ownership of the means of production rather than developmental goals, the Japanese state emphasized economic development as its preeminent goal. Key elements of a ‘developmental state’ included an elite economic bureaucracy insulated from routine political pressures and endowed with the autonomy to pursue development goals and the ‘perfection of market conforming methods of state intervention in the economy’ (Johnson, 1982: 317). These methods centered on specifying a set of goals and comparing the performance of firms with a set of external reference economies. Expanding the application of the concept of ‘developmental state’ to South Korea and Taiwan, Alice Amsden (1989) and Robert Wade (1990) explored the complexities of strategic industrial policies. Similarly, several authors have underlined the critical importance of state intervention in the emergence of Hong Kong and Singapore as off-shore manufacturing platforms (Castells et al., 1990; Appelbaum and Henderson, 1992; Bello and Rosenfeld, 1990; Schiffer, 1991; Rodan, 1989; Henderson, 1993).

If these accounts successfully dethroned the neoclassical orthodoxy, their tendency to view the state apparatus as a narrowly defined decision-making body ‘rather than as a set of complex and highly contested social relations’ (Choi, 1998: 51) tended to reify the state. Moreover, their exclusive focus on the administrative apparatus precluded a recognition of the wider political context of industrial planning and, at least implicitly, suggested that the installation of similar technocratic methods of industrial guidance will lead to a replication of East and Southeast Asian trajectories of growth in other low- and middle-income states (Friedman, 1988: 5, 30).

Additionally, though advocates of the developmental state were clear in distinguishing the relatively autonomous state structures along the Asian rimlands from other ‘newly-industrializing countries,’ once developmental state apparatuses had been installed, they tended to view these as perpetual growth machines. In particular, there was no recognition that rapid economic growth could tilt the balance of power from elite bureaucratic agencies toward other social actors as working classes became more class conscious and as corporations outgrew the need for subsidies and protection. Most notably, as Ziya Onis (1991: 122) forewarned, the ‘inability of the state elites to discipline private business in exchange for subsidies may lead to a situation where selective subsidies can easily degenerate into a major instrument of rent seeking by individual groups.’

Despite these reservations, there is widespread acknowledgment of the fundamental importance of state intervention. Strategies of intervention and the institutional structures of capital accumulation in each of these jurisdictions were, however, shaped by different configurations of relations between state bureaucracies, social classes, and multinational capital within the wider political and trading arrangements of US hegemony. The scope of dirigisme was also determined by their different endowments of land and natural resources, the maturity and technical sophistication of their domestic bourgeoisies, the competence and technical expertise of their pilot economic agencies, and the size of their domestic markets. The three US client states were able to embark on a strategy of state-led industrialization with large infusions of aid from the hegemonic power and American military procurements from Japan. Hong Kong and Singapore had neither the land, natural resource endowments nor domestic markets comparable to South Korea or Taiwan, not to speak of Japan, to embark on an ISI strategy. As British colonies, the two city-states were also not recipients of large doses of American aid. At the same time, while political conditions led to the creation of a vast state sector in Taiwan which was supported by a large number of small- and medium-scale enterprises, the Rhee administration sought to create a domestic constituency of support by fostering the emergence of a class of big capitalists. If the creation of gigantic industrial conglomerates in South Korea bore some resemblance to the revival of monopoly capital in Japan, organizational immaturity and technical inexperience characterized the former. Once again, the two city-states faced different constraints. The transformation of Hong Kong from an entrepôt to an industrial and financial center occurred under the auspices of a colonial administration that acted in many ways as the executive committee of large business interests. The small size of domestic capital in Singapore compelled the Lee administration to create favorable conditions to attract foreign investments, and auspiciously for the host government the timing could not have been better as an increase in competitive pressures in the core disposed TNCs to relocate the lower end of their manufacturing operations to off-shore locations to cut labor costs precisely when Singapore was expelled from the Malaysian Federation.

The combination of these divergences imply that though the reconstitution of postwar regimes in the US client states and in the two city-states strung along Asia's Pacific coasts were marked by the propensity of their administrations to intervene in economic affairs to a degree unparalleled outside the centrally-planned economies, the patterns of state intervention and institutional structures of capital accumulation were strikingly different in each jurisdiction. Consequently, rather than trace similarities in these patterns, this chapter will chart the confluence of the varied trajectories of class forces, external influences, political exigencies, and internal constraints on their respective patterns of state intervention and dominant forms of industrial organization.

In the context of the geopolitical ecology of the post-Second World War reconstitution of the world market under US tutelage, the timing of the adoption of an industrialization drive in each jurisdiction also had important consequences. While the Korean War provided a spur for the reconstruction of the Japanese economy, it led to the initiation of a strategy for economic development in Taiwan and reoriented developmental strategies in Hong Kong. If the war finally led to agrarian change in South Korea and laid the political conditions for its future development, the material and human devastation caused by the hostilities had consequential consequences for class formation. What the Korean War did for the rehabilitation of the Japanese economy, the Vietnam War did for the economies of Singapore, South Korea and Taiwan. Singapore, which gained independence more than a decade after the decolonization of Korea and Taiwan, confronted an entirely different set of problems and constraints – narrow domestic markets, low natural resource endowments, and a small national bourgeoisie. Daunting as these conditions were, deepening US involvement in Vietnam boosted demand for petrochemicals and other supplies from Singapore while Taiwan and South Korea became suppliers of a variety of labor-intensive products for the US military. The coincidence of increasing welfare benefits in the 1960s with the sharp rise in military expenditures occasioned by the war in Vietnam led to a pressing need for the US government to economize in the procurement of supplies – and it was cheaper to obtain supplies of requisite quality from Japan and the future ‘dragons’ than from within the United States or elsewhere (Arrighi, 1994: 341). Successive US administrations therefore tolerated the closure of markets along the Asian rimlands to American products while allowing enterprises in these economies relatively free access to US markets – a privilege not accorded to American allies elsewhere.

Most importantly, the different patterns of state intervention and industrial organization conditioned the emergence of networks of regional integration charted in subsequent chapters and were themselves reshaped in turn. The singular characteristic of the outward expansion of Japanese capital in the 1960s – being spearheaded by small- and medium-scale enterprises – was possible only due to the hierarchical structure of subcontracting networks and the organizational armature of the general trading companies. If the widespread filigree of small- and medium-scale enterprises in Taiwan could readily accommodate this transfer of less-skilled manufacturing processes from Japan, the large South Korean chaebol could more easily accommodate the transfer of heavy and chemical industries in the 1970s. At the same time, the practice of channeling huge loans to the chaebol, necessary to finance large undertakings, created a culture of high debt-equity ratios, the adverse consequences of which were brutally exposed during the economic meltdown of 1997–98. Discrimination by nationalized banks against native islanders in Taiwan fostered no similar culture of debt-financed industrialization. The greater weight of state-owned enterprises in Taiwan and Singapore, as we shall see in Chapter 6, enabled their governments to deploy macroeconomic controls more effectively even when the transborder expansion of corporate networks had rendered national industrial policies increasingly anachronistic in Japan and South Korea.

Politics in command?

If Albert Hirschman (1958: 5) was right in arguing that economic development ‘depends not so much on finding optimal combinations for given resources and factors of production as on calling forth and enlisting for development purposes resources and abilities that are hidden, scattered, or badly utilised,’ then in low- and middle-income countries, the state was best placed to formulate long-term developmental plans. Private investors in these economies, as advocates of Big Push developmental strategies had argued, were hesitant to make new investments because they were unsure whether other complementary investments necessary to sustain their investments would occur. Japan and the Four Dragons were especially well-placed to coordinate national industrial plans because they were relatively free from populist pressures.

This was most clearly evident in the case of South Korea, where rapid economic growth was initiated only after General Park's coup d’ état of 1961. If Singapore was spared a military coup, repressive measures implemented by the Lee Kuan Yew administration virtually emasculated political opposition to the PAP – not a single opposition candidate won an election to parliament between 1965 and 1981. Authoritarianism in Taiwan lasted until the lifting of martial law in 1987 while elections based on adult franchise were instituted in Hong Kong only in 1991, a few years before the British transfer of sovereignty to China (see So and May, 1993). Even in Japan, where democratic norms were observed, the iron grip of the LDP was loosened only in 1993, and for almost 40 years it was effectively a one-party state. Differences in power structures and in their relationships to state bureaucracies and social classes were important determinants of the patterns of state intervention and institutional structures of capital accumulation adopted in each jurisdiction.

In all three US client states, though land reform was a key element in the subordination of agriculture to a state-directed industrialization drive, the dynamics of agrarian change were conditioned by the relative potency of rural insurrection and by the specific institutional patterns of state power in each jurisdiction. Similar considerations also determined the corralling of labor movements in these states as well as in Hong Kong and Singapore. Finally, though domestic bourgeoisies in all five future ‘miracle’ economies were dependent on the state apparatus, the degrees of dependence varied considerably. If industrialists in Japan and Hong Kong were ‘mature’ and able to work in partnership with state bureaucrats and had a significant input into policy formation, the budding industrialists in South Korea and Taiwan were considerably more subordinate to political authority and the small commercial bourgeoisie in Singapore so insignificant that they were bypassed by the state altogether.

Japan

Spared from the ‘winnowing hand’ of the Supreme Commander Allied Powers (SCAP), the civil service had a virtual monopoly on governmental expertise, representing as they did almost all of Japan's remaining ‘political grownups’ (Pempel, 1998: 85). Indeed, far from being decimated, in the first three years of the Occupation, the civil service had grown by 84 percent over its highest wartime strength (Johnson, 1982: 44). A widespread consensus among the elite that the bureaucracy was vital to postwar recovery was encapsulated in a lead editorial in Ch ü ö k öron in August 1947:

under the present circumstances of defeat, it is impossible to return to a laissez-faire economy, and … every aspect of economic life necessarily requires an expansion of planning and control, the functions and significance of the bureaucracy are expanding with each passing day. It is not possible to imagine the dissolution of the bureaucracy in the same sense as the dissolution of the military or the zaibatsu, since the bureaucracy as a concentration of technical expertise must grow as the administrative sector broadens and becomes more complex.

(quoted in Johnson, 1982: 44)

This consensus reflected the close cooperation between the bureaucracy and the Keidanren.

The three main instruments for state intervention included control over foreign exchange and the ability to strategically target industries for development; provision of loans at preferential rates of interest, and tax concessions to lower production costs in chosen sectors; and the power to order the creation of industrial cartels and bank-based industrial groups (Johnson, 1982: 199). Control over foreign exchange, initially vested in the Foreign Exchange Control Board – and transferred in 1952 after the peace treaty, to MITI – lasted until the liberalization of trade in 1964. The establishment of a Foreign Investment Committee with oversight over all contracts involving foreign investments, acquisition of licenses and patents, conferred enormous statutory power on the ministry to regulate the import of foreign technology that lasted until the laws were rescindedin November 1979, almost thirty years after the law was approved as a temporary measure by the US occupation authorities (Johnson, 1982: 194, 217, 302; Morris-Suzuki, 1994: 168).

The imperative to ensure that adequate investment capital was available to meet the rapid expansion of US procurements occasioned by the Korean War led to a loosening of monetary controls as the Japanese government set up several banks to provide low-interest loans to industry. From the very beginnings, it was clear that the policy of ‘overloaning’ alone could not solve the capital shortage and ensure the industrial reconstruction of Japan. Since the Dodge Plan mandated balanced budgets, the Japanese government sought to overcome capital scarcity through the creation of new government-owned banks and the expansion of existing ones. The two major sources of funds for these financial institutions initially came from the US ‘counterpart funds’ – or the revenues in yen obtained from the sale of US aid that were held in a special account – and from the government-operated postal savings accounts. Once the Occupation had ended, the government empowered the Japan Development Bank to raise capital by issuing its own bonds, and sought to ensure the growth of deposits in the postal savings accounts by exempting the interest on the first ¥3 million of each account from taxes. As individuals could open accounts in multiple post offices, each of which was tax-exempt up to this ceiling, these accounts grew exponentially. Simultaneously, the government consolidated the postal savings accounts into a large investment account – the Fiscal Investment and Loan Plan (FILP) – which became, since 1953, the single most important source of finance for development, ranging from one-third to one-half of the general budget and not subject to legislative scrutiny until 1973 (Johnson, 1982: 207–8, 210; Eccleston, 1989: 49; Pempel, 1998: 67).

Between 1953 and 1961, over and above indirect support to industry through government loan guarantees, the government supplied between 19 and 38 percent of all capital. This support was especially critical in regards to strategic industries – electricity generation and power supply, ships and shipbuilding, coal, and steel – which accounted for 83 percent of disbursements by the Japan Development Bank. The elimination of risk in designated growth sectors by government guaranteed loans was reinforced by the tight controls exercised by the Finance Ministry over all interest rates, bank operations, and dividend rates. Since even permission to open new branches had to be obtained from the ministry, bank managers had to concentrate only on expanding the bank's share of loans and deposits (Johnson, 1982: 206–11).

Perhaps the most important consequence of the policy of overloaning was the creation of bank-led keiretsu (industrial conglomerates) in place of the zaibatsu based on family-owned holding companies. Capital shortages had encouraged each enterprise to establish close ties with a particular bank because even if

it did not necessarily get all the money it needed or preferential terms from its primary bank … it did get one thing it could not do without – access to capital in the first place because it was an established customer. The banks in turn became dependent upon the financial health of their heavily indebted priority industries and therefore took responsibility for them.

(Johnson, 1982: 205)

Apart from banks, industrial firms also re-established ties in the 1950s to the old zaibatsu trading companies that had been broken up by the occupation forces. With the end of the Occupation in 1952, the government repealed antimonopoly laws over the protests of small- and medium-scale businesses, and MITI actively encouraged the recomposition of prewar industrial conglomerates on a new foundation. The ministry's policy of ‘keiretsu-ization’ even led its Industrial Rationalization Council to assign industrial enterprises to a trading company (sogo shosha) if an alliance had not already been established. The ministry also used its licensing powers and control over preferential financing to reduce the numbers of trading companies from 2,800 at the end of the occupation to about 20 massive ones, each associated either with a bank-led keiretsu or a constellation of smaller firms (Johnson, 1982: 205–6; Borden, 1984: 164).

Crucially, the creation of cartels and bank-led industrial groups led to the mobilization of capital on a scale adequate to create competitive enterprises in the more technologically-sophisticated sectors. The pace of consolidation in the 1960s and 1970s was so extensive that by 1974 ‘five corporations or fewer controlled 90 percent or more of the markets in the steel, beer, nylon, acrylic, aluminum ore, automobile, and pane glass industries’ (Pempel, 1998: 94).

The combination of the policies of overloaning to strategic sectors and government facilitation of alliances between large banks, industrial firms, and trading companies led to the creation of what is often called ‘one set-ism,’ or the formation of a full complement of designated growth industries within each keiretsu group to prevent being excluded from virtually risk-free sectors. Each of the six main banks – Dai Ichi Kangyo, Fuji, Mitsubishi, Mitsui, Sanwa, and Sumitomo – for instance, established ties with a major automobile company: Isuzu, Nissan, Mitsubishi, Toyota, Daihatsu, and Mazda respectively. This led to fierce competition and to inevitable overproduction as each group sought to maximize its share of government-guaranteed loans (Johnson, 1982: 206–8; Yoshino and Lifson, 1986: 33; Morales, 1994: 100–1; Gerlach, 1989; Hamilton and Biggart, 1989: S57–8; Yonekura, 1993: 211–13).

In short, the Japanese bureaucracy and big business forged an alliance to resuscitate the economy after its wartime devastation. The bureau-cracy's control over vital foreign exchange and the mobilization of domestic savings through FLIP enabled MITI and the Ministry of Financeto strategically target specific sectors for accelerated growth by providing loans at preferential rates. If capital shortages fostered close coordination between enterprises, the bureaucracy used its financial leverage to reinforce these tendencies and create large units that could reap the economies of scale.

South Korea and Taiwan

Despite their shared heritage of Japanese colonial occupation, postwar regimes in South Korea and Taiwan were reconstituted on very different foundations, though both had very narrow bases of support and were almost entirely reliant on their bureaucracies and their coercive forces. Since the American occupation forces chose to work with the Korean Democratic Party (KDP), the most conservative faction, it provided an avenue for the privileged landed aristocracy, the yangban, to reemerge as the elite of the new order in South Korea (Cumings, 1981: 97; Amsden, 1989: 36–7). If this provided continuity in government personnel, the reconstituted Korean colonial bureaucracy was ill-equipped to promote economic development because the Rhee government lacked both the technical expertise for guided industrialization and a mass mobilizational party to establish its domination over civil society.

In contrast, as the GMD was staffed by a large cadre of experts with considerable expertise in industrial production and economic planning on the Chinese mainland, the regime's higher degree of autonomy from domestic classes in its island redoubt reinforced the strength of the state apparatus. If the GMD party-state thus had more policy instruments and options to pursue a dirigiste policy than most other states, Chiang's preoccupation with returning to the Chinese mainland meant that economic growth was assigned a low priority. Hence, though the Taiwan Production Board was established in May 1949 to stabilize the chaotic conditions following his retreat to the island, and its functions expanded early the following year with the creation of an Industrial and Financial Committee, there were no serious attempts to address the issue of economic growth before the Korean War.

When Chiang's fiction of maintaining that the GMD was the legitimate government of all of China and the consequent duplication of ministries at the ‘national’ and ‘provincial’ levels blurred lines of responsibility, the creation of two supra-ministerial agencies and the parallel bureaucratization of US aid disbursements shaped a distinctive political economy of development in Taiwan. These agencies – the Economic Stabilization Board (ESB) and Council on US Aid (CUSA) – chaired by the prime minister and composed of the key ministries and departments, provided a platform for technocrats to operate unimpeded by inter-departmental struggles for control of policy. Vested with a formidable array of powers – control over industrial development, monetary and banking policies, foreign trade and the allocation of foreign exchange, military spending, budget and taxation, agriculture and price stabilization, and utilization of US aid – the ESB was far more capable of monitoring the pulse of the economy and selectively targeting specific sectors than any comparable regulatory body in the Rhee administration. Indeed, precisely because state assets in South Korea had been sold at heavily discounted prices to political supporters of the regime, no elite bureaucratic organ was created to oversee the economic performance of firms. Hence even as ‘entrepreneurs’ were abjectly dependent on state patronage for access to aid allocations and privileged exchange and interest rates, the South Korean regime was quite unable to monitor their performance, as indicated by the widespread practice of ‘entrepreneurs’ selling their foreign currency allocations and industrial licenses to the highest bidder.

The ESB in Taiwan also emerged as the arena to reconcile differences between hardline conservative elements in the GMD hierarchy who remained hostile to the development of a powerful bourgeoisie among the indigenous islanders and US aid officials advocating the creation of a strong private sector. At the same time, as hopes for a triumphant return to the mainland receded by the end of the Korean War, Chiang's administration recognized that economic growth was a better guarantee for its long-term survival since it could dilute indigenous islanders' resentment of their political and economic domination by mainland émigr ées. Given these considerations, the regime and its US advisors sought to channel funds toward the promotion of agriculture and the creation of infrastructural projects that would encourage the broadest possible expansion of industrial production as long as the units of capital accumulation remained ‘small and until the point where its transactions involve the external world’ (Wade, 1990: 268).

The promotion of rural development assumed cardinal importance not only because the party hierarchy attributed their humiliating defeat on the Chinese mainland to the exploitation of the peasantry by the landlords with whom the party was closely identified (Gold, 1986: 68; Wade, 1990: 82, 246–8, 260), but also because technocrats could siphon agricultural surplus toward industrial development by manipulating the terms of trade against agriculture. Greater support for agriculture – the rural sector received 21.5 percent of all US aid as opposed to 15 percent for industry (Jacoby, 1966: 50–1; Cheng, 1990: 45) – consolidated the GMD party-state's hold on the small peasantry while ensuring that the native Taiwanese did not pose a credible threat to the economic dominance of the émigr ées. Support of the farm sector also satisfied the demands of US advisors that aid disbursements be more equitably shared while manipulation of the terms of trade was tantamount to a recycling of these funds to industry (Ho, 1978; Simon, 1988a: 148; Wade, 1990: 82–4).

In these conditions, the shift to project-specific grants which gave Agency for International Development (AID) officials greater control over their disbursements strengthened the tendency toward a relatively even regional distribution of funds within the island. Thus, almost two-thirds of all non-military assistance were allocated to the development of infrastructural projects – in power generation, transportation, communications, and education – which generated substantial external economies and provided the framework for the emergence of a broad base of consumer goods industries in the private sector, especially in such key sectors as plastics, rayon, glass, soda ash, and hardboard. American prodding also led the government to establish the Industrial Development and Investment Center and the China Productivity Center in 1960 to promote the growth of the private sector, convert the system of multiple exchange-rates to a single rate, and ease restrictions on foreign trade to encourage foreign private investments (Jacoby, 1966: 134–8; Ho, 1978: 195–7; Gold, 1986: 69, 76–8; Koo, 1987: 168; Haggard and Cheng, 1987: 115; Simon, 1988a: 148–50; Wade, 1990: 52–3, 202; Bello and Rosenfeld, 1990: 238).

Whereas only a few indigenous Taiwanese entrepreneurs – like the journalist Wu San-lien, who fronted the Tainan Textile Corporation for several business men from Tainan, or the five largest former landlord families who controlled the Taiwan Cement Corporation – had benefitted because of their ties to the GMD before economic conditions had stabilized in the island in the mid-1950s, these measures led to the emergence of a new breed of entrepreneurs among the indigenous islanders by the end of the decade. This was not, however, indicative of a loosening of state controls over the economy. In the first instance, though some of these entrepreneurs – such as Wang Yung-ch’ing and Lin Tin-sheng – were to become major players in their own right, public enterprises still received the bulk of aid allocations: 67 percent of the assistance to the industrial sector as a whole, as opposed to 27 percent for mixed enterprises and only 6 percent to private firms (Gold, 1986: 71–3, 82; Gold, 1988b: 189; Wade, 1990: 91; Bello and Rosenfeld, 1990: 237–40). In the second instance, government control over the banking sector ensured the party-state's control over credit allocations as well as enhancing its ability to channel domestic savings toward priority sectors, especially to enterprises controlled by the GMD. Finally, the extent of state control over domestic sources of finance did not end with government ownership of the banking sector as firms required prior approval from the Securities and Exchange Commission, an agency with wide discretionary powers, to raise capital by issuing shares.

The difference between the broad-based and increasingly autonomous and self-generating development in Taiwan, and the comparatively narrow-based and heavily aid-dependent South Korean trajectory was manifestly evident by the end of the 1950s. Within five years of the end of the Korean War, US aid officials were confident enough of the reconstruction of the Taiwanese economy to begin a gradual phase out of assistance. Despite exceptionally high rates of growth, there was no such complacency regarding South Korea. Given the generalized venality of state and business elites in South Korea, the tapering off of American assistance – after peaking at $383 million in 1957, US economic aid fell to $321 million in 1958 and to $222 million in 1959 (Woo, 1991: 46, table 3.1, 72) – precipitated a sharp decline in growth rates as it undermined the provision of subsidized credit to industry. By another measure, per capita income in South Korea in 1960 was only $62 – the same as in Japan in 1868 (in 1960 prices) – and lower than in Ghana, Senegal, Liberia, Zambia, Honduras, Nicaragua, El Salvador, and Peru (Hart-Landsberg, 1993: 26; Mason et al., 1980: 181; Cumings, 1987; Amsden, 1989: 41; cf. Wade, 1992: 277, n. 21).

Growing economic pressures added fuel to the fire of a student-led opposition to the authoritarian Rhee government, leading ultimately to the military coup on May 16, 1961. If issues of political stability and national unification had dominated the agenda of the Rhee administration, rapid economic growth dominated the agenda of the military junta, particularly since students and the urban middle-class had spearheaded the ouster of the former regime and the countryside had remained quiescent (Koo, 1987: 169). Or as Alice Amsden (1989: 49) pithily puts it: ‘If the lesson that the United States learned from the Korean upheavals was the need for stability before growth, then the military learned that causality ran in the opposite direction, from growth to stability.’

The changed priorities of the new regime entailed a restructuring of government organization and the institutional conditions for capital accu-mulation more generally. Within five months of the coup, the military government established comprehensive controls over the financial infrastructure and capital flows by re-nationalizing commercial banks, subordinating the Bank of Korea to the Ministry of Finance, and forming two new state-owned financial institutions, the Medium Industry Bank and the National Agricultural Cooperatives Federation. In addition, the functions of the Korean Development Bank were enlarged to enable it to borrow from overseas sources and to underwrite foreign loans of domestic enterprises (Kuznets, 1977: 78; Amsden, 1989: 16, 72–3; Woo, 1991: 51–2, 84). As a result, by 1970, the government controlled an astonishing 96.4 percent of the country's financial assets (Bello and Rosenfeld, 1990: 51).

While these changes had some similarities with Taiwan, there were also marked contrasts reflecting differences in the balance of class forces in the two jurisdictions and in the nature of their regimes. The most notable feature of the reconstitution of political power in South Korea after the coup was a greater centralization and concentration of power in the executive. As in Taiwan, the regime created a pilot agency, the Economic Planning Board (EPB), which controlled the national budget, foreign investments, and government guarantees of overseas loans. Its mandatory powers were augmented by the fact that its chief chaired the Council of Economic Ministers and since 1963 was also designated deputy prime minister. However, the creation of two secretariats in the Blue House – the presidential mansion – and of a Board of Audit and Inspection which reported directly to the president provided independent checks on the EPB. These measures were accompanied by a reorganization of government ministries, with the Ministry of Trade and Industry being charged with the promotion of exports and controls on imports, plans for industrial development, industrial licensing, approval of applications for investments and the designation of strategic projects and firms; the Ministry of Finance with the regulation of all financial institutions, tax assessment and collection, and foreign exchange; and the Ministry of Construction with infrastructural development (Cumings, 1987: 72; Johnson, 1987: 154; Haggard, 1990: 64–5; Hart-Landsberg, 1993: 48–50, 54).

If this reorganization of the government apparatus enabled the new regime to monitor economic activities on an almost daily basis, the creation of watchdog agencies that bypassed the elite bureaucracy underlined President Park's inability to create a mobilizational party modeled on the GMD. When the failure of his Democratic Republican Party to become a mass-based party became evident in the elections of 1963, which Park won only narrowly, organs of the state began to function as substitutes for front organizations of a mass party. Among these, the most prominent was the Korean Central Intelligence Agency which combined internal and external information-gathering capacities and vastly augmented the surveillance capabilities of the state. Similarly, the new government's decision to directly appoint all the staff of local agricultural cooperatives, who had previously been elected by members, reflected an attempt to strengthen control over rural areas since fertilizers and government credit were solely distributed through the cooperatives.

Paralleling this move, the government made it mandatory for all incorporated businesses to join one of 62 producer associations which functioned as intermediaries between the Ministry of Trade and Industry and individual firms, and acted as conduits for information. Through these associations, the government was able to negotiate price controls on strategic goods and services, grant industrial licenses, and promote targeted sectors. The new government also reorganized and strengthened the powers of the Korean Foreign Traders’ Association, which had powers to arbitrate international trade disputes, grant import and export licenses, and monitor individual firms’ compliance with government trade regulations and targets. Finally, the military disbanded all labor unions and created the state-sponsored Federation of Korean Trade Unions in which all candidates for offices required government approval (Amsden, 1989: 16–18; Haggard, 1990: 62–3; Hart-Landsberg, 1993: 51–3).

Though these measures to reinforce the regulatory capabilities of the state bore striking resemblances to the measures implemented in Taiwan, differences between the two regimes were equally striking. For one thing, the uneven economic record of South Korea made it increasingly vulnerable to American pressure. When Park's freedom to maneuver had been constrained by bad harvests and high domestic inflation, increased pressures from US aid officials compelled a reluctant regime to devalue the currency in 1964, raise domestic interest rates in 1965, and normalize diplomatic relations with Japan the same year. In these conditions, the Park administration sought to reduce its reliance on American aid by vastly expanding overseas borrowing. In contrast, rather than resorting to increased overseas borrowing when US economic aid to Taiwan was being phased out in the late 1950s and early 1960s, and ceased completely in 1965, the GMD regime eased restrictions on foreign investments, which rose from 1.7 percent of gross domestic capital formation between 1952 and 1960 to 6.9 percent between 1969 and 1974 (Simon, 1988a: 149–50).

Paradoxically, a major impetus to increased overseas borrowing by the South Korean government and enterprises was the insistence by US aid officials and International Monetary Fund advisors that domestic interest rates be raised to mobilize private savings and lessen dependence on foreign borrowing. This advice was, in turn, based on the experience of Taiwan where high interest rates had led to a rise in domestic savings from approximately 5 percent of national income in the early 1950s to more than 30 percent in the late 1970s.¹ However, in South Korea, a virtual doubling of domestic interest rates in 1965 rendered the cost of domestic borrowing more expensive than borrowing from overseas and reinforced the dominant position of the government, as it had amended to the Foreign Capital Inducement Law in 1962 to guarantee foreign loans and thereby eliminate the risk of default and exchange-rate depreciation. The shift to overseas borrowing was so pronounced that net indebtedness rose from $301 million in 1965 to $2.57 billion in 1970, with two-thirds of the loans in the late 1960s being from private sources, and the debt-equity ratio of manufacturing firms increased from 1:2 in 1966 to 3:9 in 1971. Put another way, by one estimate, without this tremendous inflow of foreign capital, South Korean production in 1971 would have been smaller by a third (Kuznets, 1977: 78–80; Jones and Il SaKong, 1980: 101; Frieden, 1987: 149; Amsden, 1989: 73; Cheng, 1990: 157; Hart-Landsberg, 1993: 58–9; Koo and Kim, 1992: 127–8).

The obverse side of the increasing resort to overseas borrowing in South Korea was an extremely restrictive foreign investment policy. Though the exhaustion of the ISI phase led the Park administration to slightly ease restrictions on foreign investments after 1962, such investments continued to be prohibited in many sectors and even in those sectors that were legally open to outside investors, government regulators routinely rejected applications not deemed to be in the ‘national interest.’ In practice, then, majority equity participation by foreign owned firms was seldom approved unless they were entirely export-oriented and the preferred option was to permit joint ventures with a local partner to transfer technology. As a result of these measures, FDI accounted for only 3.7 percent of net capital transfers to South Korea between 1967 and 1971 while the equivalent figure for Mexico was 36.6 percent, for Brazil 33.8 percent, and for Thailand 26.1 percent (Evans, 1987: 207; Amsden, 1989: 74–7; Bello and Rosenfeld, 1990: 54–5; Hart-Landsberg, 1993: 86–90).

Complementing these restrictive policies on overseas investments were stiff import tariffs which were retained even after the exhaustion of the ISI strategy. Though quantitative controls on imports were gradually eliminated, the imposition of special tariffs since 1961 to absorb the differential between domestic prices and landed cost of imports implied a steep increase in the average legal tariff. Similarly, the shift from a positive list system whereby listed goods could not be imported without explicit government permission to a negative list system when listed goods were automatically approved for import in 1967 masked a variety of practices which sharply restricted the range and volume of imports – and the share of freely importable goods in total imports declined from 55.6 percent in 1968 to 46.7 percent in 1974 and to 38.8 percent in 1978. By another measure, it has been estimated that, even as late as 1990, less than 3 percent of imports could be classified as luxury goods (Kuznets, 1977: 153–4; Krueger, 1979: 89–92; Mason et al., 1980: 128–32; Bello and Rosenfeld, 1990: 52–3; Hart-Landsberg, 1993: 37, 78–81; Haggard and Moon, 1993: 72–3).

The choice of foreign investments over increased overseas borrowing in Taiwan and the opposite in South Korea reflected fundamental differences between the two regimes. The failure of the military regime to institutionalize itself through the creation of a mass mobilizational party on the lines of the GMD led President Park to attempt to substitute economic performance for organizational capacity as the basis for political legitimacy (Cheng, 1990: 159). Once economic growth assumed priority, it was imperative that an alliance be established with the largest firms since ‘the only viable economic force happened to be the target group of leading entrepreneurial talents with their singular advantage of organization, personnel, facilities and capital resources’ (Kyoung-dong Kim quoted in Haggard and Cheng, 1987: 111).

The emerging ‘sword-won’ alliance between the junta and leading industrialists in South Korea was, however, based on an entirely different foundation from the alliance instituted under the Rhee regime. By virtue of its stranglehold over the financial infrastructure, the government was able to shape the direction of industrial production through the allocation of subsidized capital, credit guarantees, and favorable – even negative – interest rates to targeted firms and industries.² As a result, the debt-equity ratio of firms in South Korea averaged between 300 and 400 percent in the 1970s when compared with 100 to 200 percent for Brazilian and Mexican firms or 160 to 200 percent for firms in Taiwan. Moreover, by 1981, over 200 types of policy loans – targeted for specific industries at rates lower than the already highly discounted rates, and over which the banks had no control – had evolved to further promote specific types of manufacturing activity (Woo, 1991: 12). By virtue of its control over industrial licensing, the government could also reward firms entering sectors with long-fruition lags or high risks with licenses in the more lucrative sectors. Due to the higher technical requirements of targeted sectors, and the emphasis on increased exports, the regime tended to favor larger firms often controlled by political supporters: the Ssangyong group in cement rather than the more established Tongyang Corporation, for example, or the state-owned Pohang Iron and Steel Company, or the Hyundai group in shipbuilding, or the trio of Hyundai, Samsung, and Daewoo in the machine building sector (Amsden, 1989: 14–18, 73).

However, in return for privileged access to capital and industrial licenses, the government imposed stern discipline, most often related to export targets. Export-related criteria assumed preeminence in assessing enterprise performance both because continued improvement in exports provided a reliable indicator of efficiency and because tight control over the allocation of industrial licenses and cheap credit and high debt to equity ratios rendered financial indicators a poor guide. To increase market share overseas, firms were not only provided with a variety of subsidies – including lower taxes on export earnings, accelerated depreciation allowances, and duty-free imports of selected capital and intermediate goods – but were also allowed to sell products at inflated prices in domestic markets to partially off-set the costs of dumping products abroad. At the same time, the government imposed restraints on the market power of large conglomerates by negotiating price controls annually, and as late as 1986 some 110 commodities ranging from flour and sugar to automobiles and chemicals were subject to such controls (Kuznets, 1977: 156–62; Krueger, 1979: 92–9; Amsden, 1989: 17. 144–51; Bello and Rosenfeld, 1990: 52–3).

If secure access to cheap credit and a battery of incentives encouraged chaebol to pursue expansion into areas with long-fruition lags and high risks, it heightened their dependence on the government and the regime did not hesitate to dismember firms that failed to fulfill their export targets or other performance criteria without plausible excuses. Even the very largest conglomerates were not immune to such sanctions, as indicated by the experience, for instance, of the Shinjin company which had a larger share of the domestic automobile market than Hyundai in the 1960s. However, as Shinjin could not survive the oil crisis of the early 1970s and competition from Hyundai's ‘Pony,’ its credit lines were cut and the government, in its role as banker, transferred the company's assets to Daewoo Motors (Amsden, 1989: 15; Bello and Rosenfeld, 1990: 70–1; Hart-Landsberg, 1993: 69–70; see also Cumings, 1987: 74). In this context, it is significant that though the government imposed strict discipline on the chaebol and had no hesitation in dismembering and cannibalizing poor performers, business failure did not lead to unemployment as assets were simply transferred to other politically better connected conglomerates (Amsden, 1989: 15, 139–55; Eckert, 1993: 102–4).

With the rapid increase in exports and the emergence of successful entrepreneurial groups there was both a greater stratification of wealth in Taiwan and the emergence of wealthy indigenous islanders. By 1983 seven of the leading ten leading business groups were led by native Taiwanese (Bello and Rosenfeld, 1990: 238–9). Equally importantly, the predominance of small- and medium-scale firms and their spatial dispersal across the island,³ provided a relatively even regional development of industrialization on the island, with the obvious exception of the EPZs. The percentage of industrial employment in the five big cities, for instance, declined from 36.8 percent in 1966 to 22.8 percent in 1986, while it rose correspondingly in the four suburban regions from 31.9 percent to 45.8 percent. The increasing dispersal of industrial production was reflected in the decline of their unit size. While the average size of industrial plants in the five big cities declined from 26.9 employees in 1966 to 21.6 employees in 1986, it remained relatively stable in the four metropolitan regions, falling from 25.8 in 1966 to 25 in 1986, and rose from 13.5 to 21.8 in the twelve rural regions (Deyo, 1989: 20–1, 41; Amsden, 1991: table 6; see also Cheng and Gereffi, 1994: 210).

In sharp contrast, one important consequence of the strategy of ‘betting on the strong’ firms in South Korea was the progressive growth of regional imbalances. Ironically, reflecting their rural background, the military junta had initially attacked the illicit accumulation of wealth under the First Republic, and demonstrated a marked bias toward agriculture. Almost immediately after the seizure of power, they liquidated most of the farm debt and shifted the terms of trade in favor of farm produce so that rural and urban incomes were on par by 1965. However, once emphasis shifted toward rapid industrialization, the centralization and consolidation of political power in the executive and the concentration of economic power in the chaebolmeant that despite Seoul's proximity to one of the most volatile political frontiers of the Cold War, urban and industrial growth was skewed toward the capital city where more than 96 percent of all firms were headquartered. Another pole of industrial concentration, reflecting the high dependence on imports and exports and the need to lower transportation costs, has been the Pusan–Kyongsang corridor on the southeastern coast. The polarization toward these two nodes is indicated by decline of population in the Cholla provinces on the southwest seaboard, falling from some 25 percent of the population in 1949 to 12 percent in 1983, and its share of manufacturing employment showed similar decline, falling from 13.1 percent in 1958 to 5.4 percent in 1983. In contrast, the population of the Kyongsang provinces rose from 28 percent of the total in 1949 to 30 percent in 1983 and its share of manufacturing employment from 28.6 percent in 1958 to 40.1 percent in 1983 (Chon, 1992; Douglass, 1993b).

Clearly, the polarization of industrial production toward these two industrial nodes was a result of the greater emphasis placed on industrial production at the expense of the rural sector. Due to the concentration of investments in these growth poles and the predominance of large-scale industries, employment opportunities were not only disproportionately greater in Seoul and the Pusan–Kyongsang corridor, but the pattern of investments had the effect of precluding the development of small- and medium-scale industries in the rural areas. Moreover, for those without land, re-entry to the farming sector was difficult (Cheng, 1990: 160–1; Deyo, 1990: 195).

Regional imbalances, however, reflected more than merely the subordination of agriculture to the state-led industrialization drive. The Kyongsang provinces provided the political base of the military junta and 32 percent of government officials from 1961 to 1986 were recruited from these provinces. Similarly, the founders of nine of the top 20 chaebos also came from these provinces. Conversely, as the Cholla provinces, the richest agricultural zone in the country, declined in relative terms, it became the center of resistance to the military (Chon, 1992).

Briefly put, however much Taiwan and South Korea may have inherited a shared legacy of Japanese colonialism, differences in the constitution of their postwar regimes, and the adoption of different patterns and tools of state intervention led to sharp divergences in their economic structure:

The logic of the Korean approach – hierarchical, unbalanced, and command-oriented – calls for the intensive use of resources to foster a highly select and obedient business sector to carry the specific tasks the leadership may assign. The logic of the Taiwan approach – horizontal, balanced, and incentive-oriented – implies the extensive use of resources to allow a more pluralistic economy within the broad parameters delimited by the state.

(Cheng, 1990: 142)

These differences in industrial structures were to have consequential consequences over time. The greater addiction of South Korean chaebolto high debt–equity ratios and debt-financed expansion made them household names the world over as they trespassed with impunity into technologically-sophisticated sectors but made them vulnerable to the vicissitudes of short-term capital markets, as amply evident by their exposure to the economic crisis that began to unravel their corporate structures in 1997. If the smaller size of Taiwanese enterprises meant that they did not enjoy the same brand name recognition, their smaller exposure to foreign loans largely insulated them from the financial crisis that engulfed the chaebol, as we shall see in Chapter 5. Conversely, the spatial concentration of industrial production endowed workers in South Korea with much greater political power than their Taiwanese counterparts.

Hong Kong and Singapore

Dirigisme in the two city-states was based on an entirely different set of conditions. The loss of their hinterlands meant that their entrepôt role was fatally compromised and their small markets, poor endowment of natural resources, and the absence of large infusions of US aid ruled out the incubation of a domestic bourgeoisie. If Hong Kong's large expatriate industrial bourgeoisie and the migration of Chinese manufacturers from Shanghai partly ameliorated its conditions, no such relief was available to the Singaporean government. Not only did they have to confront and corral a strong left-wing movement which had no counterpart in Hong Kong, but the British naval withdrawal further exacerbated Singapore's economic plight. Governments of both city-states were therefore compelled to intervene in a manner designed to transform their territories from entrepôts to off-shore manufacturing platforms and to provide conducive conditions for foreign investments.

In Hong Kong, there was a great degree of cohesion between government and business elites as evidenced by the fact that all major legislation was circulated in draft form to the main employers associations – the Federation of Hong Kong Industries, the Chinese Manufacturers' Association, the Hong Kong General Chamber of Commerce, and the Employers' Federation of Hong Kong. While the General Chamber, dominated by the big British ‘hongs,’ even nominated one of the members to the Legislative Council, similar privileges were not accorded to the labor unions that were not represented on the Council. However, employers' associations rarely acted in concert. The sheer numbers of small firms – there were 141,708 establishments in 1981, of which 47,996 were in manufacturing – made consensus difficult while the political weakness and fragmentation of labor meant that there was little incentive to close ranks. Besides, close interrelationships between major firms and their informal ties to the colonial administration made it possible for them to safeguard their interests without recourse to the larger associations (Deyo, 1989: 14, 43–5; England and Rear, 1975: 12; Lethbridge and Ng Sek-Hong, 1993: 92–5).

While the British government appointed the governor and all senior officials in the Crown Colony, and nominated all official and unofficial members of its Executive and Legislative Councils and had the power to override local laws, local officials had almost complete autonomy in practice and the last time the parliament at Westminster overruled local laws in Hong Kong was in 1913. Apart from foreign affairs, citizenship, and landing rights for aircraft (which enabled British airlines to negotiate reciprocal rights with airlines in East and Southeast Asia), laws were enacted entirely by local officials, even though the Governor was theoretically vested with almost unlimited powers – subject only to the stipulation that he consult with the Executive Council composed of six senior officials and eleven nominated unofficial members. In practice, laws were passed by the Legislative Council. Close links between the officials, who often serve a longer term than the governor's five-year stint, and directors of major corporations led to a great deal of social cohesion among the government and business elites, as indicated by the appointment of top government officials, after their retirement from the civil service, to key executive positions in employers' associations.

By shouldering the costs of infrastructural development and subsidizing wage-goods, the colonial government provided conditions conducive to an expansion of low-cost manufacturing activities. The very success of these policies led to greater state involvement in the economic arena as the rapid rise of cheap exports from Hong Kong led to the imposition of quotas on their exports by the governments of the United Kingdom, the United States, and other high-income states from the late 1950s. Beginning with the textile industry, and subsequently expanding to an increasing number of sectors, the impositions of quotas prompted the colonial administration's Industry Department to set quotas for individual firms to ensure that Hong Kong's global quotas would be equitably shared between the major firms. However, though the government had initially allocated quotas on the basis of installed capacity, the quotas soon became tradable commodities between textile mills and garment manufacturers (Lau and Chan, 1994: 116; Miners, 1993: 110–11; Castells et al., 1990: 90–1, 119–21; Yeung, 2000: 144).

Despite protestations of non-interference in the economy, the precariousness of the small-scale industries on which its export production was based encouraged the colonial administration to provide a variety of services to these firms that they could not otherwise obtain for themselves. Thus, beginning in the mid-1960s, as other low- and middle-income states began to attract foreign investments and to abandon ISI policies for reasons more fully explored in the next chapter, the Hong Kong government established a number of public agencies – most notably, the Hong Kong Tourist Association, the Hong Kong Trade Development Council, the Hong Kong Export Credit Insurance Corporation, and the Hong Kong Productivity Council – to provide commercial intelligence, personnel training, legal and technological assistance, and insurance for high-risk ventures for small- and medium-scale enterprises. Collectively, these measures amounted to an ad hoc industrial policy until 1979 when the government formulated an explicit industrial strategy to upgrade the technological level of industries as well as to transform Hong Kong into a financial service center (Castells et al., 1990: 90–1).

The absence of a strong domestic bourgeoisie imbued dirigisme in Singapore in shades of a very different color. Rather than seeking to incubate or revive a home-grown class of entrepreneurs, government policies were fashioned to attract foreign investments to transform the island-state into an offshore manufacturing platform. Such investments also served to preclude the emergence of domestic Chinese capital as an alternate base of power as well as ensuring that Western powers had a stake in Singapore's survival. Once the militancy of labor had been curbed and steps taken to reduce the costs of reproduction through public housing schemes, the government moved to create an institutional framework to entice foreign investors. Toward this end, and given Singapore's relatively high wage structure, the government offered foreign investors a raft of incentives: reduction of taxes on the profits of exports of manufactured goods to one-tenth of the standard corporate rate; exemption of machinery, equipment, and raw materials required for industrial production from all import duties; unrestricted repatriation of profits; and generous depreciation allowances (Rodan, 1989: 87; Bello and Rosenfeld, 1990: 291–3; Haggard, 1990: 111).

As in other developmental states, the EDB determined industrial policy and aimed to target sectors of increasing technological sophistication for accelerated growth. However, rather than nurturing domestic industries, the EDB assessed applications for ‘pioneer industry status’ which enabled designated sectors to receive low-interest loans, tax holidays, and other special privileges (Haggard, 1990: 113). The institutional scaffolding for the more efficient implementation of these policies was provided by a range of specialized agencies: the Jurong Town Corporation created in June 1968 to oversee the development of industrial estates and lands; the Development Bank of Singapore (DBS) in July 1968 to provide long-term finance at low-interest rates; the International Trading Company (Intraco) in November the same year to lower procurement costs of raw materials by buying in bulk and to expand overseas markets; and the Neptune Orient Lines in January 1969 to reduce dependence on foreign shipping. Finally, not content to rely solely on private capital, the government also established 13 new public manufacturing enterprises in 1968 and eight more the following year (Rodan, 1989: 94–5).

The focus on attracting foreign investments inevitably meant that the government neglected to nurture domestic entrepreneurs and this had become starkly evident by 1976 when locally-based capital in manufacturing declined from S$123.1 million or 42.2 percent of all manufacturing investments in 1974 to S$42.8 million or 42.1 percent in 1976. Consequently, the government launched the Small Industries Finance Scheme in 1976 to advance low-interest loans for the establishment of small industries and for the diversification of existing ones. However, since protectionist measures would contravene the primary strategy of attracting large foreign investments, the intent behind small industry promotion was to create synergies between local industries and the TNCs to create fully integrated industrial sectors, especially in the electronics industry. The expansion of Japanese subcontracting networks in the 1970s limited the success of this strategy since low-cost suppliers were emerging in neighboring locations as limitations of narrow domestic markets in Malaysia and Thailand, ethnic conflicts in Malaysia, and the oil boom in Indonesia led to an opening of these economies at the same time. As these tendencies led to a less than expected degree of integration in the manufacturing sector, the Singaporean government introduced the Product Development Assistance Scheme in 1978 to boost research and development capabilities of local firms (Rodan, 1989: 124–5).

Just as different socio-political constitutions led regimes in South Korea and Taiwan to shape distinct political economies on their shared legacy of Japanese colonialism, so too did the governments of Hong Kong and Singapore shape distinct political economies on their shared legacy of British colonialism. Despite their differences though, the reintegration of these economies into the post-Second World War world market exhibits some striking parallels. The fallout of the Chinese Revolution prompted the GMD regime in Taiwan and the British colonial administration in Hong Kong to promote or subsidize small-scale manufacturing to accommodate refugees from China, as well as to broaden the regime's constituency among the indigenous islanders in Taiwan and to compensate for the Crown Colony's loss of its hinterland. A different dynamic shaped the political economy of South Korea and Singapore. Confronted by the need to quickly incubate a domestic bourgeoisie and also to bolster its support base, the Rhee administration sold off expropriated Japanese assets to its supporters. Reckoning that reliance on the big conglomerates will yield quicker results, the Park regime continued to support them even in the face of opposition from US aid officials. Without the influx of manufacturers that conditioned Hong Kong's transformation into a locus of low-cost manufacturing, the Lee administration in Singapore sought to entice large foreign investments as manufacturers in Western Europe and North America faced increasing competitive pressures in their home bases. Finally, while the resuscitation of the Japanese economy also led to the revival of large-scale enterprises, these were tied in a symbiotic relation-ship to small- and medium-sized enterprises, as we shall see in the next section.

Patterns of industrial organization

By the early 1980s, when economies strung along the Asian rimlands registered the fastest rates of growth, several analysts attributed these high growth rates to the ‘Confucian family values’ they shared, including the subordination of sectoral interests to the larger national interest, the respect for education, and the predilection for hard work. The spectacular rise of Japan was said to spring from the three ‘sacred pillars’ of its dominant pattern of employment relations – lifetime employment, seniority wage system, and enterprise unionism. It was suggested that stability of employment provided the basis for cooperative relations between workers and management while seniority wages replicated family hierarchies and by providing an array of benefits (from housing to holiday travel), enterprise-based unions reinforced the image of company-as-family. However, closer examination revealed that Japanese corporations provided long-run employment rather than lifetime employment and that this was also the case with large, well-managed corporations elsewhere. Small enterprises, which constitute the overwhelming majority of firms in Japan and else-where, were never able to provide similar conditions of employment nor were they able to provide welfare benefits on the scale of enterprise unions of large corporations (Aoki, 1987).

Nevertheless, industrial structures and employment relations in these economies evolved along lines markedly different from the normative Euro–North American patterns. Unlike in the United States, for instance, where hiring can take place at any time during an individual's career trajectory and is done in a decentralized manner, Japanese firms tend to hire at graduation through a centralized personnel department. However, while Japanese workers can more easily change jobs within the firm once they are hired, US workers tend to get locked into specific job categories. Finally, while US workers tend to be paid by skill and job classification and only secondarily by seniority, workers in major Japanese corporations tend to be paid according to seniority rather than by job classification, but seniority was never a sacrosanct principle in Japan when layoffs were concerned (Morales, 1994: 54). Yet, as briefly alluded to in Chapter 2, rather than being traditional, these employment practices were the outcome of management strategies to improve labor productivity and union attempts to improve the status of industrial workers and their working conditions in the 1950s (Gordon, 1985; Shapira, 1993: 242).

The greater job security enjoyed by workers in major Japanese corporations was also due to the maturity of the working class. Elsewhere along the rimlands, where industrial working classes were constituted after the Second World War, the ideological offensive against ‘communism’ generated by the wars in Korea and Vietnam and the very different institutional structures of accumulation conditioned employment conditions. Though workers in colonial Korea had a tradition of militancy, war and partition had so decimated the old working class that workers in the postwar era were recruited from the countryside and virtually none of them had prior experience in wage-employment (Koo, 2001). In Taiwan and Hong Kong, on the other hand, the predominance of small-scale industries led to a reconstitution of patriarchal, family-based workshops that often generated a new gender politics articulated in the idiom of familial values (Greenhalgh, 1994). While a full discussion of these patterns of employment relations and the forms of resistance they generated is outside the purview of this chapter, we seek to sketch dominant patterns of industrial organization in which workers were embedded in Japan, South Korea and Taiwan, and Hong Kong and Singapore. Complementarities and differences between industrial structures in these jurisdictions conditioned the emergence of regional networks of trade, production, and investment charted in the next two chapters. Thus, this integration of studies on their varied characteristics provides an essential background to our analysis.

Japan

When the Japanese government after the Meiji Restoration in 1868 embarked on a crash program of industrialization, they created several key institutional innovations, perhaps the most notable of which was the sogo shosha or general trading company. These were diversified enterprises with interests ranging from procurement, financing and transportation of capital goods and raw materials to the distribution and sale of finished goods. With their vast network of agencies, the sogo shosha established alliances with the zaibatsu, serving as their primary if not exclusive distributors, and fledgling industrial ventures to reap economies of scale in the purchase of raw materials. The sogo shosha's large network of procurement agencies and distribution outlets provided critical market intelligence and freed their client firms from bearing the costs associated with maintaining such systems of their own. Even after the US occupation forces began dismantling monopoly houses – including the two largest sogo shosha, the Mitsui Bussan and the Mitsubishi Shoji – senior executives of the former companies were aware of the benefits that would accrue from pooling their resources and as soon as the Occupation ended, the Japanese government facilitated the reconsolidation of the sogo shosha. By virtue of its diversified activities a sogo shosha could spread costs over a number of transactions and thereby lower unit costs for its member firms. Its ready access to raw materials, markets, and information could reduce risks and monitor price movements far better than any one of its client firms, no matter how large, could do by themselves. Banks also preferred lending directly to sogo shosha as it freed them from processing a large number of small loans while a sogo shosha's intimate knowledge of market conditions and a firm's creditworthiness reduced risks of default (Yoshino and Lifson, 1986). Yet, since the client firms were independent entities, they were not necessarily tied to one sogo shosha and may indeed sell their products to outside firms or to clients of rival sogo shosha.⁴ At the same time, each sogo shosha often included rival firms who competed with each other. Hence, as Michael Yoshino and Thomas Lifson (1986: 47) note, a sogo shosha product system 'is best understood not as a rigid body whose constituent units are mechanically linked into a tightly balanced system but as a constellation of firms active at various stages of a complex production process.' Nevertheless, the sogo shosha represented an enormous concentration of economic power with the top ten controlling between 50 and 60 percent of Japanese foreign trade and some 20 percent of its domestic wholesale trade in the 1960s (Pempel, 1998: 70).

After the premature termination of the program for zaibatsu dissolution, postwar restructuring of industrial organization in Japan was marked by two distinct forms of industrial conglomerates – the intermarket groups or kigyo shudan (for example, Dai Ichi Kangyo, Fuyo, Mitsubishi, Mitsui, Sanwa, and Sumitomo) and independent industrial groups or kaisha (for example, Hitachi, Industrial Bank of Japan, Matsushita, Nippon Steel, Nissan, Tokai Bank, Tokyu, Toshiba-IHI, Toyota, and Seibu). As the name suggests, intermarket groups were horizontal coalitions of firms built around financial institutions and sogo shoshas, while the independent industrial and financial groups represented vertically integrated firms in one or more sectors built around a large parent company. In both cases, control over affiliated firms was established through interlocking stock ownership, though the degree of control exerted tended to be stronger among intermarket groups than among firms within the independent groups. Though the six major horizonal keiretsu represented only 0.1 percent of all the companies in Japan, they accounted for almost 25 percent of the total value of shares in the Tokyo Stock Exchange (Eccleston, 1989: 40–1; Hamilton and Biggart, 1989: S57–9; Glasmeier and Sugiura, 1991: 399–401; Pempel, 1998: 70).

Even if horizontal linkages between financial institutions, industrial firms, and commercial organizations through interlocking shareholding, lender–borrower, buyer–seller, and director relationships provided the institutional base for the reconstruction of the Japanese economy, the scalar magnitude of US military procurements at the start of the Korean war was so extensive that major firms sought to utilize idle productive capacities of parts manufacturers by sourcing out components (Smitka, 1991: 6–10, 53–78; Eccleston, 1989: 35; Glasmeier and Sugiura, 1991: 399). As wage workers constituted only 39.3 percent of the working population at the onset of the Korean War (Itoh, 1990: 145), this strategy was seen as a temporary expedient especially since the length of the war-induced boom was uncertain. Large firms, however, soon abandoned their efforts to recreate their prewar structure of vertically integrated production systems as the advantages offered by stable subcontracting arrangements rapidly became evident.

From these beginnings, as several theorists associated with such ungainly coinages as ‘Fujitsuism,’ ‘Toyotism,’ and ‘global Japanization’ have underscored, subcontracting arrangements in Japan evolved in a manner that contrasts sharply with similar practices elsewhere. First, as indicated by Table 3.1, though small- and medium-scale firms as a percentage of all manufacturing enterprises in Japan are approximately the same as in other core states, they account for a much higher percentage of employment, shipments, and value added to production. Correspondingly, small and medium scale enterprises in Japan were highly dependent on a few large firms and in the early 1980s over 90 percent of all subcontractors in the textile sector were found to have sold the bulk of their output to three firms or less (Eccleston, 1989: 30–1; Glasmeier and Sugiura, 1991: 401; Pempel, 1998: 71).

Second, though Japanese manufacturers may depend on fewer primary subcontractors than vertically-integrated corporations in other high-income states, subcontracting networks are more extensive in Japan due to the highly stratified nature of these networks. Thus, whereas a typical Japanese automotive maker has 170 primary subcontracting (ichiji shitauke) firms (producing machinery such as robots, jigs, and large body panels; subassemblies like engines and seats; and major body parts such as brakes), the latter depend on 4,700 secondary (niji shitauke) subcontractors (supplying dies, metal work, small body parts, and such single components as brake linings) who, in turn, depend on 31,600 tertiary subcontractors (sanji shitauke). Below this layer are the large numbers of households where women fabricate minor parts – metal stampings of brand names, tiny electronic components, etc. (Sheard, 1983: 56; Cusumano, 1985: 250–3; Hill, 1989: 464–6; Fujita and Hill, 1993: 180–2; Morales, 1994: 108–9; Arrighi et al., 1993: 51). A government survey on the state of industry in 1976 revealed that more than 90 percent of firms employing 50 workers or more, and more than 75 percent of firms with 10 to 40 workers farmed out work to subcontractors. What was still more remarkable was that almost 50 percent of firms employing four to nine workers and one-sixth of firms with one to three workers issued subcontracting work (Sheard, 1983: 60; Cusumano, 1985: 192; Kenney and Florida, 1988: 136).

Third, instead of purchasing components under short-term contracts from a large number of suppliers, as was the norm for vertically-integrated US corporations until recently, Japanese firms built long-term relationships with a small number of primary suppliers (Smitka, 1991: 4–10, 58–88, 175–89). The continuity of these affiliative relationships enabled major Japanese manufacturers to circumvent the high transaction costs, irregular deliveries, and poor quality of parts that plague outsourcing arrangements in other core states, and in pre-Second World War Japan itself (Morris-Suzuki, 1994: 154–5; Glasmeier and Sugiura, 1991: 399). This was very different from the arm's-length, market-based, short-term contracts with independent companies that was the norm in the United States and Western Europe.

By farming out production to a large number of suppliers who have no direct relationship with each other, mass manufacturers in the United States were often confronted with incompatible parts. In contrast, by relying on a smaller number of primary subcontractors, Japanese manufacturers tended to merely specify performance requirements and leave engineering decisions to their first-tier suppliers who were responsible for an entire component (Womack et al., 1990: 60–1, 142, 146–7).

This crucial difference between cooperative subcontracting relation-ships constructed by major Japanese manufacturers in the 1950s and 1960s, often idealized as a familial relationship in contrast to the fundamentally antagonistic buyer–seller relationships in arm's-length arrangements, enabled major manufacturers to reduce transaction costs, lower inventories through the vaunted kanban (‘just-in-time’) system, and monitor quality control. Originated by Toyota, but soon pervasive throughout corporate Japan, kanban denotes a card attached to each batch of parts supplied by a subcontractor to a firm higher up the production chain.⁵ When the first part is taken from the consignment by an assembly worker, the card signals that it is time for the subcontractor to prepare another consignment. Due to the close coordination that resulted from frequent deliveries of small consignments, the ‘parent’ firm regulated the production rhythms in subcontracting firms and effectively trans-formed them to spatial extensions of the assembly line. The delivery of components in small batches also facilitated continuous and multiple improvements in product design, component quality, and to the production process itself. Since continuous adjustment of deliveries was crucial to the smooth operation of the system, subcontracting firms tended to be located in close proximity to the ‘parent’ firm – and in some cases, as when suppliers were located within the premises of the larger firm, the distinction between firms itself became tenuous – and this led to further reductions in transaction costs and minimized possible disruptions and delays of deliveries. In contrast, since mass-manufacturing corporations in the United States sourced components from parts manufacturers on the basis of cost, their suppliers were dispersed much further afield. Consequently, to avoid possible disruptions of supply, they were forced to stockpile parts and bear much higher overheads. Larger stockpiles in the ‘just-in-case’ system also made it unfeasible to make continuous design modifications (Sheard, 1983: 53–4, 61–3; Hill, 1989: 469–70; Eccleston, 1989: 36; Gwynne, 1991: 63; Fujita and Hill, 1993: 183–4).

First-tier subcontractors, in turn, were themselves large firms and their size and technological sophistication implied that, despite their close relations with major manufacturers, they were not totally dependent on the ‘parent’ enterprise. Though Toyota holds about 22 percent of shares in Nippon Denso, its supplier of electrical products, for example, the company also supplies parts to some of Toyota's competitors, while a large number of the 203 first-layer suppliers for Subaru were also major subcontractors to Nissan. By virtue of its size, Nippon Denso was also able to implement new innovations and Toyota's first quality-control system was, in fact, modeled on the one pioneered by its electrical parts supplier (Sheard, 1983: 56–9; Cusumano, 1985: 250–3; Kenney and Florida, 1988: 136–7; Womack et al., 1990: 61).

Extensive inter-group shareholding between firms in the multi-layered subcontracting chain – the share of cross-ownership in 1974 ranged from about 30 percent in Mitsubishi (30.6 percent) and Sumitomo (27.9 percent) to less than 20 percent in Mitsui (17.4 percent), Fuyo (17.4 percent) and Dai Ichi Kangyo (16 percent) – insulated their managements from unwelcome takeover bids, and thus encouraged them to focus on long-term performance rather than short-term profits (Kenney and Florida, 1988: 153, n. 78; Eccleston, 1989: 31–3; Womack et al., 1990: 61). Employees of the smaller firms sometimes labor alongside workers in the ‘parent’ firms, and even take responsibility for complete sections. Primary subcontractors similarly transfer equipment at nominal cost and provide worker training and financial assistance to secondary subcontractors, who in turn replicate these patterns right down to the level of the household. To share ideas and promote greater cooperation, subcontractors of major firms are also organized into kyoryokukai, or associations (Cusumano, 1985: 252; Kenney and Florida, 1988: 137; Odaka Konosuke et al., 1988; Womack et al., 1990: 153; Morales, 1994: 101).

Correspondingly, the evolution of a complex, hierarchically integrated network of subcontracting arrangements meant that major Japanese manufacturers were not burdened with the rigidity of long-term, large-scale fixed capital investments. Whereas large investments in standardized, mass-production techniques predisposed both American enterprises and their suppliers against basic design changes, the lower investments entailed by flexible small batch production processes in the kanban system enabled Japanese manufacturers to market a greater diversity of models and thereby create and dominate several market niches (Friedman, 1983: 361).

Extensive government support underpinned the evolution of this pervasive network of subcontracting arrangements. In the early 1950s, MITI set up an office of Small and Medium Enterprises and the Japan Small Enterprises Corporation to coordinate a nation-wide policy of support to these firms and to assist them to upgrade their technology, to help establish alliances with larger firms, to provide market intelligence and to provide low-interest loans. Additionally, in each prefecture a bank was designated to assess the requirements of small businesses and to help them develop business plans. These regional banks were also instrumental in creating non-profit associations called shokotai to provide training in business administration, tax planning, and labor relations for small enterprises. The difference in the level of government support to small enterprises in Japan and in the United States is vividly illustrated by the fact that the Japanese government had $66 billion in outstanding loans to the these firms in 1982 compared with the $3.8 billion that the US Small Business Administration loaned out, despite the American population being twice as large (Glasmeier and Sugiura, 1991: 406–7; Fujita and Hill, 1993).

Compared with the replication of the technostructures of US corporations in post-Second World War Europe discussed in Chapter 1, the mode of corporate governance in Japan was strikingly different. The more flexible contractual arrangements between ‘parent’ firms and their suppliers, organized in a multi-layered subcontracting network, enabled major Japanese manufacturers to drastically reduce the huge corporate bureaucracies associated with multi-unit technostructures of US corporations. In contrast to vertically-intergrated production structures, where the size and complexity of administrative hierarchies compromised effectiveness, the looser coordination of production processes by the sogo shoshas reduced uncertainty for its client firms without burdening them with an increasingly cumbersome bureaucratic structure (Yoshino and Lifson, 1986: 39–47). ‘Just-in-time’ production processes enabled Japanese firms to lower inventories, implement better quality controls and faster and continuous design improvements, and provide a greater diversity of products at lower cost. Subcontracting work arrangements also enabled manufacturers to segment the workforce and thereby undermine the unity of the labor movement and sap its strength.

South Korea and Taiwan

Despite the expropriation of Japanese corporate assets within their jurisdictions by postcolonial regimes in South Korea and Taiwan, differences in socio-political conditions meant that the corporate structures that evolved from these shared foundations were strikingly different. Though in their size and in the diversity of their product range, the South Korean chaebol bore more than a passing resemblance to the Japanese keiretsu, the former were invariably family-held enterprises. Whereas family ownership of firms in other countries was diluted due to the need to raise capital for growth, the ready availability of credit guarantees and subsidies from the South Korean government enabled the chaebol to preserve their family ownership structure (Amsden, 1989: 128; Mason et al., 1980: 276–7). Consequently, decision-making tended to be highly centralized within these.

enterprise structures and the major subsidiaries and many minor ones were headed by family members (Amsden, 1989: 151–5; Whitley, 1990: 57–8; Orru et al., 1991; Hamilton and Biggart, 1989: S58–9)

The chaebol also represented a marked concentration of economic power, virtually unparalleled elsewhere, as indicated by Table 3.2. In comparative terms, the largest 100 manufacturing firms in South Korea accounted for 40.6 percent of all manufacturing sales in 1970, 43.6 percent in 1972, 44.9 percent in 1977 and 46.9 percent in 1982. Corresponding figures for the top 100 manufacturing firms in Japan were 28.4 percent in 1975 and 27.3 percent in 1980 (Koo, 1987: 176; Amsden, 1989: 115–37; Gereffi, 1990: 95–7; Bello and Rosenfeld, 1990: 63–75; Koo and Kim, 1992: 136–7; Douglass, 1993b: 158; Hart-Landsberg, 1993: 63–5; Eckert, 1993: 122).

The product range of these chaebol was equally remarkable. Though like industrial conglomerates elsewhere the chaebol initially diversified into related businesses either because of the advantages of scope provided by a multi-divisional enterprise structure, or because local suppliers were unavailable, or because a market of sufficient size had not developed to permit specialization, very quickly the chaebol diversified into several unconnected businesses. The Lucky-Goldstar group, for example, diversified from consumer electronics to petrochemicals while Hyundai was prominent in the automobile and shipbuilding sectors. In contrast to the experience in the United States and other core states where diversification into unrelated fields occurred during periods of industrial decline or market saturation (Chandler, 1977: 481), diversification in South Korea occurred during phases of high-speed economic growth perhaps because of the chaebol's ‘relative lack of technical expertise to build upon in related products or in higher quality product niches. Their widely diversified structures complemented their strategy to compete at the bottom end of many markets’ (Amsden, 1989: 151).

Given that the wellsprings for the remarkable concentration of economic power were government patronage and the immaturity of technical development, unlike the Japanese keiretsu, the chaebol did not develop an extensive subcontracting network. The increasing scale of industrial production is vividly illustrated by the growth of large manufacturing establishments: whereas firms with more than four and less than 20 employees had accounted for 29 percent of total employment and 19 percent of the value added in manufacturing in 1963, their share had dropped to 16 percent of employment and 7 percent of value-added in 1972. Correspondingly the share of establishments with over 200 employees had risen from 34 percent of employment and 47 percent of value-added in 1963 to 55 percent and 72 percent respectively in 1972 (Kuznets, 1977: 165).

The socio-political basis of the GMD party-state – particularly its links with small farmers, labor, and small industrialists – precluded a similar alliance between state and big capital. In particular, as the land reform legislation imposed rigid ceilings on land ownership, increases in farm incomes were invariably channeled into the manufacturing sector, especially since high positions in government or in enterprises run by the GMD were effectively closed to the native Taiwanese in an ethnically-bifurcated society (Greenhalgh, 1994: 766). This is indicated, among other things, by the decline in the percentage of full-time farming in rural households from 47.6 percent in 1960 to 31.9 percent five years later (Cheng and Gereffi, 1994: 209, n. 29). At the same time, since credit allocations from nationalized banks tended to discriminate against indigenous islanders in favor of mainland émigrées and GMD-controlled enterprises, native Taiwanese entrepreneurs were compelled to rely on kin and community networks to mobilize capital and labor and it has been estimated that by the early 1970s over 82 percent of the capital raised by small and medium entrepreneurs came from the informal money market (Bello and Rosenfeld, 1990: 242). Typical of the emergent pattern of industrial organization was, hence, a bifurcated structure with a few relatively large firms dominated by the mainland émigrées at the top and a large network of family firms (jiazquqiye) and business groups (jituanqiye) at the bottom, as indicated by Table 3.3.

The rapid proliferation of these small industries – Ramon Myers (1984: 516) reports, for instance, that Wang's success in establishing Formosa Plastics led to such an extraordinary spurt of imitators that the plastics industry grew at an annual rate of 45 percent between 1957 and 1971 and while only 100 small firms processed products supplied by Formosa Plastics in 1957, over 1,300 small units bought from plastic suppliers in 1970 – led to a saturation of Taiwan's narrow domestic markets in small consumer goods, precisely when US aid was tapering out. This situation was exacerbated by prevalent inheritance patterns which led to a division of land between all male heirs who often set up small manufacturing operations to compensate for their smaller land holdings. In these conditions, a

liberalization of controls over foreign trade and encouraging foreign investments through tax incentives – most notably the 1962 Statute for the Encouragement of Technical Cooperation and the creation of the Kaohsiung Export Processing Zone in 1965, the year US economic assistance ended – was an attempt to break out of the economic bottleneck through increasing exports.

Though the initial response from overseas investors was relatively lukewarm, as indicated by Figure 3.1, once General Instruments established a bonded electronics factory near Taipei in 1964 foreign investments soared rapidly, rising from US$18 million in 1963 to $248.8 million in 1973 (Gold, 1988b: table 9.5). Underlying this impressive rise in overseas investments was the search by American enterprises for low-cost labor – the wages of a skilled worker in Taiwan in 1972 were $73 a month when they were $102 in South Korea, $122 in Hong Kong, $183 in Singapore, and $272 in Japan – to reverse the inroads being made into their domestic markets by cheap Japanese imports at a time when peripheral and semiperipheral states in Latin America and elsewhere were beginning to impose restrictions on the operations of TNCs. At the same time, Japanese corporations also increased their investments to recapture their market shares in the United States being lost to off-shore plants of US enterprises, as well as to evade quota restrictions being placed on Japanese exports to core states. Prior colonial ties further predisposed Japanese enterprises to

locate a disproportionate percentage of their offshore investments in Taiwan. Finally, persecution in Malaysia and Indonesia and additional incentives offered to overseas Chinese entrepreneurs by Chiang's administration as a part of its ideological claim to represent the legitimate government of China was responsible for the rise in investments by Chinese émigrées (Gold, 1986: 79–81; Gold, 1988b: 190–8; Bello and Rosenfeld, 1990: 245).

Apart from the availability of a literate, docile and low-wage workforce, the absence of irksome health and safety regulations, and the provision of a raft of tax incentives, the existence of a dense network of family-based small enterprises also provided a conducive environment for foreign investments. This was particularly true in the case of the small- and medium-scale enterprises which spearheaded Japanese investments as they not only lacked the resources to make large investments but also because the history of Japanese colonialism made it politically expedient to establish joint ventures with a local partner. Hence, the average size of the 349 cases of Japanese investments in Taiwan between 1964 and 1970 was $235,000 compared with an average of $1.6 million for the 123 cases of US investments during the same period and $230,000 for the 587 cases of investments by overseas Chinese investments. Even though the larger size of American investments indicated that US corporations tended to set up wholly-owned operations to export products back to their domestic markets, they often contracted out the production of intermediate components to the small family firms to further hold down production costs. Consequently, while the number of manufacturing firms in Taiwan increased by 250 percent between 1966 and 1976, the number of employees per firm increased by only 29 percent. In contrast, during the same period the number of manufacturing establishments in South Korea increased by only 10 percent, though the number of employees per firm doubled (Wade, 1990: 67).

Since holding companies are legally prohibited in Taiwan, these figures may underestimate the size of enterprises as census data do not include business groups. Nevertheless, they clearly indicate that increased foreign investments were compatible with Taiwan's bifurcated industrial structure. Encouragement of foreign investments did not amount to a reversal of an ISI strategy, as the government in its role as gatekeeper for the national economy evaluated proposals in terms of their ability to open up new markets, create new exports, transfer technology, and deepen internal economic relationships between component suppliers and final assemblers. The government also sought to safeguard local producers by insisting on foreign investors exporting their products and not competing with local enterprises in the domestic market. The power to restrict or expand imports gave the government a powerful leverage against domestic producers and overseas companies (Gold, 1986: 85; Wade, 1990: 150–1; Hart-Landsberg, 1993: 41–2).

The general protection extended to local producers not only safe-guarded the GMD party-state's domination of the commanding heights of the economy but also vastly expanded the range of opportunities for the jiazquqiye and the jituanqiye. The combination of deliberate denial of access to credit to these firms with the provision of attractive tax incentives for exporters actively encouraged the creation of OEM (original equip-ment manufacturer) arrangements with Japanese and American TNCs (Hart-Landsberg, 1993: 41). Cheap products by small- and medium-scale Taiwanese firms, marketed under such well-known brand-names as Sears, J. C. Penny, Hewlett-Packard, IBM, Sony, Sharp, Mitsubishi, National, Westinghouse, Wilson Sporting Goods, and Schwinn Bicycles, were so pervasive that the executive of an American transplant operation told researchers that ‘You really can't consider Taiwan an exporting nation. Taiwan is simply a collection of international subcontractors for the American market’ (quoted in Bello and Rosenfeld, 1990: 243).

A telling testimony to the sheer pervasiveness of subcontracting relationships is provided by the lack of brand name recognition for Taiwanese products, despite the fact that on a per capita basis Taiwan's trade surplus in 1984 was twice that of Japan (Hamilton and Biggart, 1989: S54; Cheng and Gereffi, 1994: 211, n. 35). If OEM arrangements meant that more small enterprises participated in export trade in Taiwan than in other states – between 1978 and 1985 it was estimated that small- and medium-scale firms were responsible for over 60 percent of exports from Taiwan (Wade, 1990: 70; Cheng and Gereffi, 1994: 209) – these arrangements also ensured that the benefits accruing to these firms from increased exports were disproportionately small. According to one estimate, Japanese sogo shosha control some 50 to 70 percent of Taiwan's exports to the United States. By another reckoning, the export price of many commodities from Taiwan is said to be only 10 to 20 percent of their retail price in the United States (Bello and Rosenfeld, 1990: 243–4).

As might be expected, relations between local firms and their overseas partners exhibited considerable variation, though in most cases there was a distinct asymmetry in these relationships that worked against the former. In fields such as textiles and food processing where local firms dominated, they relied on the TNCs for access to export markets. In petrochemicals and microelectronics when the local collaborator was typically a GMD-controlled enterprise, TNCs supplied technology and capital while the local firm contributed capital and its downstream plants purchased the output. In other sectors, especially in consumer electronics, the relation-ship was far more complex since partnership arrangements with Japanese investors routinely tied the Taiwanese firms in a web of contractual obligations which compelled them to purchase raw materials from, and sell the completed products to, their Japanese partner (Gold, 1986: 82–3; Gold, 1988b: 189–90). If the government proved relatively ineffective in preventing Japanese firms from implementing very restrictive subcontracting arrangements on Taiwanese firms, it also turned a blind eye on local firms counterfeiting foreign labels and manufacturing substandard items (Simon, 1988b: 216; Wade, 1990: 268)!

The minimal vigilance the government exercised over small companies was also indicative of the much lower capacities of the government to monitor economic activities when compared with South Korea. As many analysts have noted, while the government formulated economic plans it had few procedures to monitor their implementation and one study found that there were an estimated 480,000 peddlers in 1984, with some 20,000 in Taipei alone, and that two-thirds of them were unregulated. Another study indicated that only 2,687 of the 16,000 construction firms were legally registered in 1984 (Cheng and Gereffi, 1994: 210). In fact, the government's most potent instrument – its control over industrial licensing – was effective only on the larger companies. Similarly, Robert Wade (1990: 162–4) argued that Taiwan's equivalent of the subsidy in South Korea was the postdated check – when a loan is negotiated the borrower gives the lender a postdated check covering the principal and interest – which was a poor substitute as far as monitoring economic performance is concerned.

Compelled to rely on kin and community networks, the dominant characteristic of small manufacturing and service firms in Taiwan was that each enterprise was owned by a single proprietor or family and functionally linked together in a system of satellite factories or weixing gongchang to produce a single product. These inter-enterprise networks, sometimes between members of the same family who owned separate firms and some-times between unrelated proprietors, based on noncontractual agreements regarding quantity and price of their products, were the foundation stones of the structure of industrial and service organizations in Taiwan. Precisely because small family firms – merely extended households in many cases – had neither access to large resources nor the ability to forecast conditions in export markets, to diversify risks successful businesses tended to expand horizontally rather than to integrate vertically in interrelated businesses (Hamilton and Biggart, 1989: S66–7; Cheng and Gereffi, 1994: 209–10).

Unlike the unified management structure of South Korean chaebol, Taiwanese jituanqiye or business groups tended to be associations of single-unit enterprises managed by close relatives and the owners often held several managerial positions in different firms within the group. A detailed survey of the 96 largest jituanqiye in 1985 indicated that 59 percent were owned and controlled by family groups and a further 38 percent represented partnerships between unrelated individuals which could be expected to become family-based organizations in the next generation (Hamilton and Biggart, 1989: S66–7).

In short, though industrialization in both South Korea and Taiwan had been initiated by Japanese colonial authorities in the early twentieth century, the political economy of their post-colonial reconstitution led to the installation of very distinct structures of accumulation in these two jurisdictions. In South Korea, recognizing that only economic growth could guarantee political stability, the Park regime funneled enormous resources to the chaebol and used government control over finances to discipline the conglomerates. Different political conditions in Taiwan dictated a different strategy: a more decentralized and dispersed pattern of manufacturing based on small- and medium-sized enterprises with the government retaining control over the commanding heights of the economy. Similarly, though the two city-states of Hong Kong and Singapore had functioned as entrepôts in the British Empire, their transformation into off-shore manufacturing platforms in the 1950s and 1960s entailed very different patterns of industrial organization.

Hong Kong and Singapore

Hong Kong's economy was dominated by large expatriate holding companies, often associated with the old trading companies of the nineteenth century – Butterfield & Swire; Jardine, Matheson & Co. Ltd; Hong Kong & Whampoa Dock Ltd; Wheelock Marden Co. Ltd; Hutchison International Ltd; and China Light & Power Co. Ltd, among others. Not only were they the largest employers in the colony in the immediate aftermath of the Second World War, but through a system of interlocking directorships, they formed a powerful political bloc and were well represented in the Executive and Legislative Councils. When the loss of the Chinese hinterland deprived them of their markets and sources of supply, they shifted their activities toward manufacturing even though they continued to maintain their traditional interests in shipping, shipbuilding, public utilities, and trade. The Chinese Revolution also led to an exodus of manufacturers from Shanghai to Hong Kong, as we have already seen, and they were central to the transformation of Hong Kong from a colonial entrepôt to an export-oriented industrial platform. They dominated the textiles sector but by the 1960s had also diversified into garments, plastics, knit-ware, and enamelware. The overwhelming majority of workers in the manufacturing sector, however, continued to be employed by the Cantonese firms. Though many of them were established only after the Second World War, they had spread to all branches of industry. Finally, the development of Hong Kong as an export-oriented industrial platform also led to the establishment of large US manufacturing concerns, chiefly in the electronics sector. Factories established by US corporations tended to be larger in scale than those established by expatriate British firms or by Shanghainese or Cantonese firms. Though wages in US-owned electronic assembly firms tended to be higher than similar operations in other export-oriented Asian locations, their wage levels were lower than in the plastics, garments, and wig industries in Hong Kong and accordingly suffered from high labor turnover (England, 1971: 224–30).

Hong Kong's small domestic market and the lack of protection to industries had three major consequences for its economic structure. First, growing competitive pressures in labor-intensive industries encouraged a slash-and-burn attitude as entrepreneurs sought to reap quick profits. In his survey of industrial relations, Joe England (1971: 216–17) observed that in the 1960s, the large note-issuing banks who provided the bulk of funds for industries expected their loans to turn over within five years, and hence most investors did not plan beyond this term. As a result, bank credit to the manufacturing sector in Hong Kong consistently remained much lower than in the other dragon economies and even declined as it increased in the others (Yeung, 2000: 137). Second, as other low-cost sites of export-oriented industrialization serially emerged they exerted a corresponding pressure on wages and hence wage levels in low-skilled, laborintensive manufacturing sectors were determined more by the productivity of firms than by the bargaining power of labor. This was reflected by the higher average hours worked per loom: 8,160 hours per year in Hong Kong in 1965 compared with 6,401 hours in South Korea, 4,864 in Taiwan, 4,786 in Japan, and 4,149 in Thailand (England, 1971: 212; Owen, 1971: 148, table 4.5). Third, the continuing domination of the commercial and financial factions of capital meant that services contributed over 60 percent of the colony's GDP as early as 1970 (Yeung, 2000: 126).

In manufacturing, extensive government subsidies of basic wage goods and the massive public housing program, in the context of the influx of refugees – the 1981 census indicated, for instance, that net immigration accounted for 58 percent of the population increase between 1976 and 1981 (Cheng and Gereffi, 1994: 201) – created propitious conditions for the development of an extensive network of subcontracting arrangements, especially since the colonial administration routinely ignored violations of environmental and safety regulations in the housing estates. Thus, as indicated by Table 3.4, the average number of employees per industrial establishment

declined from 52.5 in 1951 to 20.0 in 1981. By another measure, while the bulk of manufactures from Hong Kong were exported, only 11.1 percent of manufacturing establishments received orders directly from overseas and import-export houses were responsible for almost 57 percent of overseas orders in the mid-1970s. Moreover, some 75 percent of these import–export houses were themselves small-scale firms (Sit et al., 1979).

Unlike in the case of Japan, however, small firms in Hong Kong were generally not exclusively tied to a single large manufacturer or trading company and according to one study only 14.7 percent of small enterprises considered themselves to be satellites of larger firms (Sit et al., 1979). Instead, the main commercial intermediaries – except in the case of the electronics sector where production and procurement were more tightly organized by the TNCs – placed orders with smaller firms, based on their assessment of market demands in high-income states. The latter firms, in turn, subcontracted out part-processes to still smaller firms on the basis of personal and family connections. The smaller firms were responsible for the purchase of their raw materials and components directly from wholesalers, and often the subcontracting processes took place within the premises of a larger firm as when, for instance, the ironing of garments were subcontracted to an individual or a family unit who performed the operations in the larger firm on piece rates. Similarly, it was not unusual for a group of workers of an enterprise to try and outbid their own firm for an order by undertaking to provide it at a lower cost. While the lack of hierarchical relationships organized by large enterprise groups as in Japan may make the subcontracting networks more flexible, it also led to greater insecurity. Although no systematic data on business failures exist, a study in 1984 found that 57.1 percent of the firms surveyed had been in existence for less than four years and that only 22 percent of the firms had been in operation for longer than 10 years (Castells et al., 1990: 66–74; see also England, 1971: 212).

Given the precariousness of small enterprises in Hong Kong, most of which were family-owned and managed, it was not uncommon for them to produce goods in high demand – such as toys and other plastics, gloves, and garments – to order as subcontractors and then invest in an entirely different line of production according to new orders from larger firms. The inherent instability of small firms meant that though a hard core of technicians was essential, other workers tended to be treated as casual labor. Correspondingly, workers had little loyalty to any firm or industry and changed jobs frequently (England, 1971: 223–4).

The prevalence of widespread subcontracting arrangements and the proliferation of small workshops in Hong Kong stood in sharp contrast to Singapore where, reflecting the government's strategy of relying on TNCs for investment, subcontracting arrangements with locally-owned firms were minimal. Though the electronics industry, to take one instance, was estimated to have created over 80,000 new jobs between 1968 and 1981 and sourced over 50 percent of its material inputs locally, it signally failed to generate backward linkages with Singaporean-owned firms. Since the smaller scale and telescoped industrial experience of Singaporean manufacturers meant that they were generally not competitive in the early years of industrialization in price, product range, technological sophistication, and time-critical delivery schedules, locally-sourced inputs were routinely procured from foreign-owned subsidiaries in the island state. Additionally, as only industrial ventures with over $1 million in investments were eligible for tax breaks conferred on ‘pioneer industries,’ small domestic entrepreneurs were ineligible for most of the incentives offered to the TNCs. The absence of protectionist measures also meant that Singaporean firms had virtually no insulation against adverse market conditions and some 38 percent of all wholly-owned Singaporean firms established since 1961 had declared bankruptcy by 1978 (Lim and Pang Eng Fong, 1986: 54–5, 92; Bello and Rosenfeld, 1990: 294–5).

Conversely, the stock of foreign equity investment grew by almost ten times in the eleven years between 1970 to 1981 – from $1.7 billion to $16.8 billion. While inflows of foreign investment to the island state began in earnest only after the failure of its merger with Malaysia in 1965, within four years foreign investment represented 54.2 percent of all fixed assets in manufacturing. This increase in foreign investments was reflected in a sharp rise in the value of exports from $349.2 million in 1965 to $1265.3 million by 1969, or a rise of 262 percent. At the same time, Malaysia's share of Singapore's manufactured exports fell from 53.7 percent to 15.4 percent. However, this also led to a dramatic falling off of Japanese investments – except in the shipbuilding and repair sectors – as these ventures had been designed to penetrate Southeast Asian markets rather than to export to Western Europe and North America (Bello and Rosenfeld, 1990: 293; Rodan, 1989: 99–103).

United States-based capital swiftly replaced the Japanese as leading investors in Singapore's manufacturing sector, with US investments rising from $27 million in 1967 to $131 million in 1969. Japanese investment, in contrast, grew by just $9 million between 1965 and 1969. US firms also tended to be much larger in size, with an average employment size in 1981 of 363 workers per establishment compared with 243 for Japanese firms, 226 for Swiss firms, and 164 for firms owned by Hong Kong based capital. Between 1965 and 1969, US companies were followed by British- and Dutch-based capital as leading investors, though these investments were primarily directed toward the petroleum sector (Rodan, 1989: 103; Lim and Pang Eng Fong, 1986: 54–8).

The magnitude of foreign investment in Singapore meant that it dominated all sectors of industry, unlike the case with Hong Kong and Taiwan where foreign investments were concentrated in specific sectors. This is indicated by Table 3.5 which depicts how foreign capital, which had been relatively insignificant in 1959, had come to dominate Singapore's economy by 1973 (Haggard and Cheng, 1987: 98–9).

Table 3.5 The role of foreign paid-up capital in Singapore's industry, 1959 and 1973

	Sectoral share of total paid-up capital (%)			Foreign proportion (%)
	1959	1973		1959	1973
Food	32	15		13	52
Textiles	1	11		53	89
Chemicals	3	9		19	84
Petroleum and products	0	19		0	100
Metals	11	9		54	51
Machinery	5	10		43	49
Electrical products	3	8		95	74
Scientific and photographic equipment	1	4		62	94
Others	44	15		11	83
Total	100	100		21	76

Source: Haggard and Chang (1987: 99, table 8).

The dominance of foreign firms continued unabated over the next decade as the Singaporean government sought to increase the quality of labor to counteract competition for overseas investments from other low-wage economies. Though low wages initially attracted US corporations, the rapid growth of industrial production virtually eliminated unemployment in the island by 1970. As this coincided with the opening up of Malaysia – where wage levels were lower – to foreign investments, the Singapore government sought to upgrade its nascent industrial structure. It withdrew ‘pioneer industry’ status from pure assembly operations in the electronics industry in 1970 and centralized wage determination in a tripartite National Wages Council (NWC), created in 1972. The following year, the NWC tried to force firms to adopt more capital- and technology-intensive processes by raising wage levels. These steps not only discouraged further inflows of foreign capital to labor-intensive sectors especially but also continued to undermine the Singaporean bourgeoisie unable to enter into higher value-added production. By the mid-1980s, transnational corporations accounted for over 70 percent of Singapore's industrial production, 50 percent of employment and 82 percent of exports (Rodan, 1989: 103–9, 124; Haggard and Cheng, 1987: 119; Bello and Rosenfeld, 1990: 293).

To recapitulate, though the dislocations caused by the Second World War and the Japanese defeat, the Korean War, the marginalization of prewar elites, and bipolar rivalry between the United States and the Soviet Union combined to confer greater relative autonomy on governments along Asia's rimlands, it also made it incumbent on them to seek new bases of legitimacy. Seeking to inoculate their subject populations from the appeals of revolutionary socialism by increasing their material prosperity, these governments adopted a variety of strategies – shaped by their domestic political coalitions, external alliances, resource endowments, size and maturity of their domestic bourgeoisie, and the organizational capabilities of their bureaucratic apparatus. In Japan, the largest and most powerful of these states, the bureaucracy – which did not suffer a debilitating purge – emerged as a key actor and established an alliance with the big bourgeoisie to strategically target sectors of increasing technological sophistication for accelerated development by providing loans at preferential rates of interest, tax concessions, and the creation of bank-led industrial conglomerates and cartels to reap economies of scale. The establishment of a conquest state in Taiwan as the GMD suppressed the indigenous islanders meant that no similar alliance between the bureaucracy and big business could be established in the island state. The greater organizational capabilities of the GMD state, and the provision of substantial assistance from the United States in pursuit of its Cold War aims, led to the creation of a massive state sector. Dependence on US aid, however, meant that American pressure led to the provision of assistance for the development of a broad-based, small-scale consumer goods industry in the private sector. Without a comparable cadre of bureaucrats with planning and technical expertise and a mass party, Syngman Rhee's inability to promote broad-based economic development ultimately led to General Park's coup d'état in 1961. Recognizing that only economic growth could guarantee political stability, the military junta established an alliance with the big bourgeoisie and established comprehensive controls over the financial sector by re-nationalizing all commercial banks. Simultaneously, attempting to lessen the regime's dependence on the US for aid, the government vastly expanded its borrowings overseas, thereby laying the foundations both for the spectacular growth of South Korean corporations and for their equally spectacular meltdown during the economic crisis of 1997–98, as we shall see in Chapter 5. When Hong Kong was confronted with the loss of its hinterland, the great degree of cohesion between government and business elites led to the government shouldering the costs of infrastructural development and subsidizing basic wage goods in the colony's transition from an entrepôt to an low-wage manufacturing platform. Finally, the absence of a strong domestic industrial bourgeoisie in Singapore led the government to develop the island state's infrastructure and offer foreign investors a variety of incentives as well as to create a significant public sector.

These strategies of government intervention, in turn, shaped patterns of industrial organization and employment relations. The premature termination of the zaibatsu dissolution program and vast increases in American procurements led to significant transformations in Japanese industrial organization and labor relations. Unsure of how long the Korean War-led boom would last, major Japanese corporations farmed out work to a variety of subcontractors. This was eventually institutionalized as a hierarchically organized, multi-layered subcontracting network which enabled corporate Japan to respond quickly to market changes and make incremental quality improvements during the production process, as well as to divide the labor market into segments with differential pay and benefit rates. The sogo shosha provided an organizational umbrella, moreover, to coordinate the activities of many small- and medium-sized firms within these networks.

If the size of the South Korean chaebol invited comparisons with the Japanese keiretsu, the former tended to be family-held enterprises and did not initially resort to extensive outsourcing of components. In contrast, the marginalization of native islanders in Taiwan and obstacles to them accessing capital from the nationalized banking structure led to the emergence of a large network of small enterprises based on relations of kinship and neighborhood. While the large, expatriate business houses dominated Hong Kong's economy, the influx of refugees from China and extensive government subsidies of basic wage goods and a massive public housing program led to the development of a dense network of small, family-based enterprises. Unlike in Japan, however, they were not organized within hierarchical, multi-layered subcontracting networks and were hence more precarious. Finally, the Singapore government's strategy of betting on foreign investments meant that linkages between large, vertically-integrated enterprises and domestic firms were minimal.

These complementary industrial structures led to the greater regional integration of production along the Asian rimlands. As production costs rose in Japan, the sogo sosha enabled small- and medium-scale enterprises engaged in low-skilled production processes to transfer their operations to neighboring locations – especially to Taiwan and Hong Kong where the emerging network of small-scale units provided a propitious environment. The imposition of export quotas on textiles and other commodities from Japan, and later from the other ‘dragon’ economies, by high-income states in North America and Western Europe also triggered a trans-border expansion of production networks as firms sought alternate sites of production. Greater Japanese concern with dependence on imported raw materials and environmental pollution and the consequent move to cleaner, higher value-added industries led to the progressive transfer of heavy and chemical industries to South Korea, where access to large amounts of overseas capital enabled the chaebol to enter these fields. Singapore, meanwhile, continued to be a recipient of large doses of foreign investments. We turn to these issues in the next chapter.