Foreign Investment and the Middle Income Trap in Southeast Asia

This article focuses on the post-colonial catch-up by Southeast Asian nations with developed countries. The article offers an analysis of the nature and causes of the middle income trap in Southeast Asia. It discusses various interpretations of this concept, concluding with the dichotomy between laissez-faire and interventionist development strategies. Empirical evidence is provided from the automotive industry in Malaysia and Thailand. Two rival explanations of the lack of strong interventionist policies in Southeast Asia are given, one stressing the weakness of political pressure on national governments, the other linked up with historical patterns of ethnic specialization and division. The argument draws on secondary sources and reflects on implications for the study of Indonesian economic history in the colonial era.


Introduction
'It is better to regard foreign capital as a necessary evil, and the sooner a country gets rid of it, the better, since the country can enjoy more economic benefits. For example, Japan made little use of foreign capital (meaning direct investment).' (Yoshihara Kunio, The rise of ersatz capitalism in South-East Asia, 1988: 123). In the last decades of the twentieth century, Southeast Asia became a byword for development success and an acknowledged part of the 'East Asian Miracle' (World Bank, 1993). From the 1960s onward, Malaysia, Thailand, Indonesia, and (later) Vietnam all achieved rates of both aggregate growth and poverty reduction which were among the highest in human history. The spectacular gains of this period were won mainly by two complementary means. The first was an agricultural revolution, based on irrigation, seedfertilizer technologies, and rural infrastructure, which improved the productivity and profitability of tens of millions of small farms. The second was a boom in labor-intensive, export-oriented industrial production, often funded by foreign investment, which provided jobs for millions of low-skilled workers. Besides the direct beneficiaries of these two engines of inclusive growth, millions more found employment in the -mostly informal -service sector that boomed as an indirect result of agricultural and industrial expansion. The role of the developmental state in this period was to provide conditions of economic freedom, to organize the large but technically straightforward public investments needed for smallholder agricultural development, and to supply the macroeconomic stability and outwardoriented trade policies needed to stimulate private industrial investment (Henley, 2015).
By the turn of the century, however, the low-hanging fruit that could be plucked by these means was nearing exhaustion. Rising wages and growing competition from poorer countries limited the scope for further laborintensive industrial growth, while diminishing returns and an increasingly urban population reduced the impact and utility of further agricultural development. Only two countries in the region -the tiny oil sultanate of Brunei and the city-state and international seaport of Singapore, both exceptional cases by any standards -had meanwhile succeeded in raising the incomes of their citizens to developed country levels. For the rest of Southeast Asia, despite all that had been achieved in absolute terms, the goal of following Taiwan and South Korea into the ranks of the world's rich nations remained a remote dream. Between 1960 and 2010 the third richest country in the region, Malaysia, increased its per capita GDP from about 10 percent of that of the USA to about 20 percent, and the fourth richest, Thailand, from four percent to ten percent (Hill, 2014: 5). At this rate, actually catching up with the developed countries might take centuries. Small wonder that policymakers talk increasingly of a 'middle-income trap', and that economists are inclined to demote Southeast Asia's development record to the status of 'halfway miracle' (Booth, 2016).
In this article I outline an -inevitably simplified and incompleteargument about the nature and causes of the middle income trap in Southeast Asia. The argument is based entirely on secondary sources and owes particular debts to the ever-inspiring work of Yoshihara Kunio, and to Joe Studwell's lucid, gritty accounts of Asian business life and lives. Nevertheless I believe that the synthesis as a whole has not been presented before in this form. The article is structured as follows. The first section following this introduction sketches competing interpretations of the trap and suggests that the economically 'heterodox view, which contrasts the laissez-faire development strategies of contemporary Southeast Asia with South Korea's history of active, interventionist, and successful industrial policy, provides a promising starting point. The next section summarizes, with particular reference to the automotive industry in Malaysia and Thailand, empirical evidence that without such intervention, manufacturing enterprises in developing countries tend to become enmeshed in ultimately debilitating dependency relationships with foreign capital. The two subsequent sections deal respectively with two useful types of explanation for the absence or ineffectiveness of interventionist industrial policy in Southeast Asia, one emphasizing the relative weakness of the political and existential pressures on Southeast Asian governments to succeed in making their countries industrial powers, the other highlighting the adverse political and ideological consequences of Southeast Asia's longstanding pattern of ethnic commercial specialization and internal societal division. A concluding section recapitulates the argument and reflects on its implications for the study of Indonesian economic history during the colonial era.

The Middle-Income Trap: Diagnoses and Prescriptions
Prescriptions for escaping the middle-income trap, as we shall see, vary dramatically. However there is agreement across the ideological spectrum that a large part of the challenge has to do with mastery of technology and innovation (Gill & Kharas, 2015;Wade, 2016). Whereas firms in middleincome countries use and sometimes copy imported technology and ideas, high-income countries are home to enterprises which create and own new technology and ideas, and which have much power to determine who, if anybody, gets to share them, and at what price. Although famous Northeast Asian industrial companies like Samsung and Sony take advantage of Southeast Asia's low labor costs to manufacture extensively there for the global market, Southeast Asia has no such global manufacturing brands of its own. Incubating innovative, market-leading, wealth-creating firms, analysts agree, is much more difficult than attracting them to one's shores.
Innovative firms are not built overnight. The examples of Hyundai, Samsung, Sony, Taiwan Semiconductor Manufacturing Company, and now China International Marine Containers, Infosys, Lenovo, Wipro, and others show how lengthy and resource intensive the process can be. No Malaysian firm -or for that matter no firm in Indonesia, the Philippines, or Thailand -has as yet embarked on the path to achieving competitiveness that is based on innovation (Yusuf & Nabeshima, 2009: 216.) As a consequence of their lack of indigenous technological capacity, the economies of the middle-income Southeast Asian countries 'tend to be dualistic, with foreign firms dominating high-tech exports as assemblers, exhibiting few linkages to domestic producers of intermediate and capital goods' (Doner, 2009: 9). Domestic capital and enterprise, meanwhile, concentrates on property, services, protected national consumer markets, and extractive industries, where competitive pressures and technological demands are lower.
Where diagnoses of the middle-income trap part company is on the question of what governments can and should do in order to remedy this situation. What might be called the orthodox school of thought calls for provision of more and better education; for improvement of the entrepreneurial environment through deregulation and strengthening of the rule of law, including intellectual property rights; and (sometimes) for state support for industrial research and development (Kharas & Kohli, 2011;Tran, 2013). These prescriptions, however, are arguably based more on a particular interpretation of Western historical experience than on empirical observation of the processes at work in the contemporary developing world. As a glance at the Philippines, at one time by far the most educated country in Southeast Asia, readily suggests, the causal relationship between education and development performance is actually a weak one at best (Chang, 2011). Among developing countries, correlations between economic growth and conventional 'good governance' indicators such as legal security are also surprisingly weak (Khan, 2007). In the late twentieth century, corruption in business and politics was no less a feature of rapidly industrializing South Korea than it was of the economically stagnant Philippines (Kang, 2002).
The alternative or heterodox view is that active, interventionist industrial policies are necessary to create an innovative and internationally competitive manufacturing sector (Studwell, 2013;Wade, 2008). This argument takes its cue from the experiences of the only two (large) countries which can unreservedly be said to have made the transition from underdeveloped to developed status since the mid-twentieth century, Taiwan and (especially) South Korea (Amsden, 1989;Wade, 1990). Both of these countries followed (broadly speaking) capitalist development paths, and both owed their initial success partly to infrastructure-based, smallholder-friendly agricultural and rural development policies comparable to those pursued in Southeast Asia. Their industrial strategies, however, differed markedly from those of Southeast Asia. Instead of relying on free markets, their governments restricted foreign investment while actively directing domestic investment into industries and activities judged to have developmental utility. Policy instruments here included domestic market protection, selective subsidies, and credit from state-owned commercial banks. Access to these benefits, however, was made strictly conditional on rapid success in export markets, and firms which did not meet export targets were allowed to fail.
Debate over the causes of the Northeast/Southeast Asia development divergence continues, and the industrial policy explanation is doubtless not the whole story (Lin & Chang, 2009;Chu, 2016). The interventionist model fits South Korea more completely than it fits Taiwan, where heavy industry and large corporations were less specifically targeted by development planners than in the Korean case (Dollar & Sokoloff, 1994). Comparisons between South Korean and Southeast Asian experiences nevertheless seem to validate some of the points made in heterodox interpretations of the Korean success story. In particular, such comparisons suggest that post-war Korean political leaders were right to suspect that a strong foreign manufacturing presence in the country would not ultimately benefit Korean industry. In what follows this point is illustrated with particular reference to the comparative development of the automobile industry, by most accounts a key element of the Northeast Asian (or at least, Korean and Japanese) success stories, in South Korea and in Southeast Asia.

Dependent Versus Independent Industrial Development
Three decades ago, at the height of Southeast Asia's economic progress and poverty reduction, Yoshihara Kunio warned in a provocative book entitled The rise of ersatz capitalism in South-East Asia that when it came to the big business sector, all was not as promising as it seemed in the 'miracle' economies of the region. Borrowing his terminology partly from critical literature on the political economy of Latin America, Yoshihara argued that although the Southeast Asian economies were no longer dominated by foreign capital, and although domestic capitalists were active in manufacturing as well as agriculture, construction and services, manufacturing companies remained entirely dependent on foreign firms for their technology. Yoshihara wrote: 'Southeast Asian industrial capital cannot act as the vanguard of economic development because it does not have export capability. [...] This is because large industrial capitalists are comprador capitalists (acting as the agents of foreign manufacturers in their own countries), or they depend on foreign technology in a broader sense, or they are not efficient enough to compete in the international market.' (Yoshihara, 1988: 3) Nowhere was the comprador or 'ersatz' character of Southeast Asian capitalism clearer, according to Yoshihara, than in the motor vehicle industry. In this developmentally important sector, the relationship between local manufacturers and their foreign 'partners' was such that they were structurally prevented from acquiring any skills and technologies which might enable them to develop their own export capacity. The result was a permanent and debilitating pattern of dependent development and 'technologyless industrialization'.
'South-East Asian capitalists [...] are engaged in motor vehicle and motor vehicle parts production. Their activities appear quite technologically sophisticated, and they are sometimes hailed as industrial pioneers; but the fact is that they are Japanese compradors. [...] Most of the major components are supplied by Japanese companies, and some which are produced locally are made under the technical supervision of the Japanese companies. Technologically [...] they are almost 100 per cent dependent on their Japanese licensers, and, under the present set-up, it would be impossible for them to become technologically independent and start exporting their products. Their technological dependency is not temporary but, being structural, semi-permanent.' (Yoshihara, 1988, 111-112) The rise of ersatz capitalism was quite sceptically received at the time, with critics questioning the 'static quality of the analysis' (Robison, 1989: 121) and pointing out that all developing countries, including Japan, 'have begun from a position of technological backwardness' (Booth, 1991: 162). With hindsight, however, it is clear that Yoshihara's pessimistic vision was prescient.
At the beginning of the new century a survey of the automotive industry in Indonesia, Malaysia and Thailand confirmed that while the local content of motor vehicles produced in all three countries was increasing, 'crucial technologies and skills remain jealously guarded by foreign partners in the production of both vehicles and component parts' (Abbott, 2003: 149). 'The existence of such situations of dependence [...] raises questions about the long-term ability of the industry sectors [...] to mature and compete in more value-added production. Of particular concern [...] is that the failure to upgrade their skills-base and move into higher valueadded production is resulting in their low-wage competitive advantage being eroded by new competitors such as China.' (Abbott, 2003: 151) In Thailand, the region's leading motor vehicle producer, the position of local suppliers in the industry actually weakened after 2000 as Toyota, Isuzu and the other Japanese companies which dominated the sector increasingly relied on Japanese parts manufacturers operating in Thailand, rather than on Thai partners. Taking continuing advantage of relatively low labor costs, assembly processes meanwhile remained labor-intensive, with output per worker much lower than in richer countries (Busser, 2008). In Indonesia, attempts by local manufacturers to move up the 'Global Value Chain' (GVC) of the automobile industry likewise seem to have been impeded, rather than assisted, by accelerating foreign direct investment (FDI) in the sector. 'The Indonesian automotive industry is growing rapidly, but there are still several reasons for concern. First, domestic investment in the automotive and parts industries has been extremely low compared with the trend of FDI [...]. Within the automotive GVC [...] the upgrading opportunities of local parts-supplying firms -particularly those firms looking to climb the tiers of the subcontracting system [...] may be hindered by competition from incoming foreign firms [...]. This problem was confirmed by our interviews with executive officers of the local parts industry [...].' (Natsuda, Otsuka & Thoburn, 2015: 62) Dependence on foreign technology has constrained the development of indigenous motor vehicle design and manufacture even in Malaysia, where a sustained attempt has been made to develop a truly national automotive industry under the auspices of the government-linked Proton and Perodua companies. Despite its nationalist ambitions, Proton began in the 1980s as a joint venture with the Mitsubishi Motor Company (MMC) and its first production model was effectively a Mitsubishi Lancer, rebranded as the Proton Saga. The plan was to acquire, via the partnership with Mitsubishi, the skills and technologies Proton would need to design and manufacture exportable vehicles on an autonomous basis. This, however, proved easier said than done -at least partly because it was not in Mitsubishi's interest to allow it to happen.
'MMC was able to: sell the joint venture superannuated production equipment; slow down the pace of content localisation; keep prices of its own components far above world market prices; and build a car which Japanese managers knew would not meet safety standards for export to developed countries. When Mitsubishi was forced to localise components, it frequently did so through Japanese joint ventures which kept most technological learning out of Malaysian hands.' (Studwell, 2013: 120) Studwell, the author of this account, contrasts Proton's experience with that of South Korea's Hyundai Motor Company (HMC), which chose to acquire new technology by hiring foreign experts on an individual and temporary basis rather than by entering into partnerships with foreign companies. 'A [...] key to HMC's manufacturing success was its ability to obtain foreign techology in such a way that the firm learned skills -and eventually learned how to originate its own technology -without becoming dependent on foreign multinationals. In the automotive industry, developing country car makers around the world have found it almost impossible to gain technological independence after getting into equity joint ventures with global auto firms. In joint ventures, it is too easy and too comfortable for the local entrant to become dependent Foreign Investment and the Middle Income Trap in Southeast Asia Vol. 14 No. 1 April 2018 on drip-fed technology from a foreign partner, while multinationals have no interest in helping local firms export.' (Studwell, 2013: 102) Today Hyundai is the world's third largest motor vehicle manufacturer, with sales in almost every country. Proton does not make the top forty, and has almost no export sales. In June 2017 its parent company, DRB-Hicom, sold a 49.9 per cent stake in Proton to a Chinese concern, Geely. Thailand, meanwhile, exports two million vehicles each year under many international brand names, but still has no producer of its own, remaining essentially a lowcost offshore assembly platform for overseas manufacturers. Indonesia too has followed a dependent path, and exports a growing but still much smaller number of locally assembled Japanese and Korean models.
We may of course question the extent to which the relational problems outlined above -rather than, for instance, inferior education and training, or lack of managerial ambition -are responsible for the absence of globally successful Southeast Asian car producers. How much this absence ultimately has to do with the persistent economic gap between Southeast and Northeast Asia is even more debatable. What is nevertheless clear is that 'dependent development' is not just an outdated neo-Marxist theory of Latin American underdevelopment, but a real, tangible problem for specific enterprises even in dynamic East Asia -or rather, in those parts of East Asia which have been described as caught in a 'middle-income trap'. Moreover the pattern of persistent technological dependency on Multinational Corporations (MNCs), as a World Bank report from 2009 emphasizes, is not limited to the motor vehicle industry. 'Indonesia, Malaysia and Thailand [...] have all acquired considerable manufacturing capabilities, most notably in electronic and electrical engineering, textiles, and the automotive industry. [...] However, innovation -product or process -remains mainly a preserve of the MNCs; indigenous firms do very little innovation. More disquieting is the sparseness of backward links from MNC operations, which would signify progressive industrial deepening [..]. This lack of backward links means that domestic value added in manufacturing remains low.' (Yusuf & Nabeshima, 2009: 9-10) Malaysia's foreign-owned high-technology industries are evocatively characterized by Wade as 'cathedrals in the desert', transplanted marvels lacking any substantial integration into the surrounding domestic economy (Wade, 2016: 472).
In studies of the Southeast Asian auto industry, even authors not directly engaged in Northeast Asian comparisons conclude that strong state intervention is necessary to enable domestic manufacturers to move up the value chain and avoid either permanent defeat by, or permanent subordination to, more experienced, technologically advanced, and capitalrich foreign competitors (Busser, 2008: 42;Natsuda & Thoburn, 2013: 434).
The historically indicated best practice would appear to be that of South Korea, where foreign competition was largely excluded from the domestic market, but local entrepreneurs at the same time assisted by the state to enter the industry, and forced by politically imposed export requirements to acquire the technologies and skills they needed for international competitiveness and expansion. Why have Southeast Asian states not followed this path, rather than settling for industrial dependency?

Historical Origins of Independent Industrial Development: Regime Vulnerability
Perhaps the most coherent explanation found in the academic literature for the Northeast/Southeast Asian industrial development divergence is based on the idea that it was the 'systemic vulnerability' of postwar political regimes in South Korea and Taiwan that effectively forced them to take the difficult steps needed to follow the independent rather than the dependent path (Doner, 2009;Doner, Ritchie and Slater, 2005). Faced with real threats both of internal revolution and of foreign invasion, and lacking any abundant natural resources to provide a ready-made economic basis for strong state power and national defense, their governments concluded that they had no alternative to rapid industrialization if they and the states they controlled were to survive. Under these circumstances, the creation of an autonomous and internationally competitive industrial sector became an urgent national project. 'When a nation's survival is at stake', wrote South Korean president Park Chung Hee, 'politics, economy, culture, everything should be organized and mobilized for that single purpose' (Park, 1979: 130). Japan, in a longer historical frame, can be included in the same model, faced as it was first with the threat of colonial conquest, then with the challenge of proving itself as an imperialist latecomer in a Eurocentric world, and finally with the task of recovering, under the shadow both of its communist enemies and of its new American protectors, from the disastrous defeat of 1945. The key to East Asian industrialization, in this perspective, was 'the East Asian setting of revolutionary nationalism -not a garden variety nationalism but one that grew from war and imperialism' (Woo-Cummings, 1999: 7).
Nationalism and revolution played important roles in Southeast Asian economic history too, in particular when the threat of revolutionary communism inspired national leaders to respond to rural discontent by implementing pro-poor, pro-rural development strategies (Henley, 2015). But while an internal revolutionary threat was often present, a sustained military threat from outside the state usually was not. Unlike their northern neighbors, some Southeast Asian countries had the further comfort of being relatively rich in natural resources, which gave their rulers easier ways of funding their own and their citizens' welfare than trying to create autonomous industrial strength. These circumstances arguably held Southeast Asian governments back from advancing beyond agricultural growth and labor-intensive, dependent industrialization into the more difficult arena of state-led industrial development.
In any developing country domestic capital, left to its own devices, tends to be too myopic and risk-averse to make long-term industrial investments on the necessary scale; it can only be induced to do so by means of protection and subsidy (Studwell, 2007;2013). Once these supports are in place, however, only very strong collective political will can prevent such a cosy, mutually beneficial relationship from developing between politicians and industrialists that it becomes impossible to expose maturing industries to the vital test of international competition. '[T]he case of the Korean chaebol -or the Japanese zaibatsu -is not like what one finds in the Southeast Asian countries: it is not a case of indiscriminate 'crony capitalism'. What one had in Southeast Asia was not a nationalistic mobilization for export led growth, as in Japan and Korea, but a kind of "protection ring". Indonesia was the worst case [...].' (Woo-Cummings, 1999: 19) It is the difficulty of transcending this 'crony capitalism' under all but the most severe conditions of external threat which, according to the 'systemic vulnerability' model, explains Southeast Asia's 'tendency to cede manufacturing issues to foreign producers' (Doner, 2009: 260).
This type of explanation for Southeast Asian industrial retardation is not without weaknesses. Thailand, for instance, was not only subject -just like Japan -to a threat of European conquest in the colonial period, but also to a renewed possibility of invasion (as well as domestic insurrection) during the second half of the twentieth century, when it was a front-line state in the Cold War and 'most political leaders in Thailand [...] perceived a military threat from communist Vietnam' (Warr, 1993: 19). Whether Thailand can usefully be classed as a 'resource-rich' country is also doubtful, since it lacks the major oil and gas deposits that are generally thought of as the most important reason to classify Indonesia and Malaysia as such. The claim that Southeast Asian governments have lacked serious commitment to state-led industrialization is questionable too, particularly in the case of Malaysia where long-serving prime minister Mahathir Mohamad was personally responsible for founding the company charged with the National Car Project, Proton, which he alternately led or advised for more than three decades from 1983 to 2016. Mahathir's failure to make Proton sufficiently independent of foreign partners, and to subject it to adequate 'export discipline', seems to have been caused as much by his misunderstanding of the Japanese success stories that inspired it as by any personal interests which he or his political allies had in the company's undisturbed domination of the domestic market (Studwell, 2013: 106-110).
Nevertheless, it is not my intention to contest the usefulness of the vulnerability model, which clearly does identify important forces behind the observed divergence between Southeast and Northeast Asian industrial development strategies. Rather, the following part of the present article aims to complement that model's focus on twentieth-century high politics by highlighting another circumstance, more deeply rooted in history, which also helps to explain the divergence.

Historical Origins of Dependent Industrial Development: Ethnic Capitalism
Korea and Taiwan, like Japan, emerged from the colonial era as relatively homogenous societies to which whatever capitalist elements were present ultimately belonged, just like the rest of the population. In colonial Southeast Asia, by contrast, business and commerce were largely the preserve of alien minorities. These included not only Europeans, who -like the Japanese in Korea and Taiwan -mostly departed following decolonization, but also millions of Chinese and South Asians who had been encouraged to migrate to the region by colonial governments, and whose descendants remained there in the postcolonial period as ethnically distinct communities of 'pariah entrepreneurs' (Riggs, 1966: 249-254). The full reasons for this contrast are a matter of controversy, and more will be said about them in the conclusion. But whatever the historical processes involved, the undisputed outcome is that 'the dominant commercial role of the ethnic Chinese immigrant minority in Southeast Asia has no parallel in Northeast Asia' (Lim 1996: 94).
The Southeast Asian pattern of ethnic business specialization seems to have had profound consequences for the political economy of the region's postcolonial states. Ample evidence, as noted, suggests that the goal of nondependent industrial development is most effectively pursued as an urgent national project in which all elements of society take part and in which the role of big business, albeit acting under strong state discipline, is particularly crucial.
'The developmental state in East Asia has always been a paradise for big business, and unlike the northern European 'welfare states', the protective gaze in Northeast Asia has never been down, toward the downtrodden, but rather upward, toward the privileged, to help big business compete more vigorously in the global marketplace -with its legitimacy resting in the eternal invocation of nationalism.' (Woo-Cummings, 1999: 30) In reality the Northeast Asian states, particularly during the initial agricultural phase of their development, were perhaps not as indifferent to the 'downtrodden' as this account suggests. Nevertheless it is true that in the industrial phase, rich capitalists were the most direct beneficiaries of a development strategy which still had to be cast and understood as a national endeavor serving the common interest. In countries where the capitalists mostly belong to an envied and unpopular ethnic minority, to the point that their very membership in the national community was in doubt, such an endeavor was bound to be problematic.
The developmental consequences of this dilemma are clearest in the case of Malaysia, where national politics have always revolved around the relationship between the indigenous (bumiputera) Malay majority, which is politically dominant, and the ethnic Chinese and Indian minorities, which dominate the private sector. No development policy, consequently, can be blind to the need to redress economic disparities between Malays and non-Malays. Malaysia's enduring historical stance of 'exceptional openness toward the international economy' (Drabble, 2000: 262), including foreign direct investment, may itself have been inspired in the postcolonial period by a tacit preference among Malays for any kind of development that does not further enrich -and potentially empower -Chinese Malaysians, even if it enriches foreigners instead (Jesudason, 1989: 167-168). What is certain is that the political impossibility of embracing experienced non-Malay entrepreneurs as partners in Mahathir's national industrialization drive (Bowie, 1991: 132-135) was a serious practical problem for state-backed industrial projects like Proton.
'Mahathir mixed up industrial policy with affirmative action. He came to power promising to raise up the indigenous bumiputera population. In so doing, he painted himself into a racial corner where he decided he could not use Malaysia's mostly ethnic Chinese and Tamil established entrepreneurs to run his new heavy industrial investments. Instead, he tried to implement effective industrial policy and create a new generation of Malay entrepreneurs at the same time. This was always going to be difficult.' (Studwell, 2013: 107-108) The predictable failure of Proton vehicles to achieve export competitiveness in the 1980s pushed Malaysia inexorably back toward the strategy, by then already tried and tested, of relying on foreign direct investment, particularly in the electronics sector, as the main driver of its export industries (Bowie, 1991: 142-152).
In Indonesia, comparably, state-sponsored industrial endeavors -notably in aviation -labored under the disadvantage of having to work through inefficient state-owned enterprises and with inexperienced managers, rather than trying to build on existing commercial institutions and managerial expertise in the private sector (Goldstein, 2002;McKendrick, 1992). In relation to the Philippines, Yoshihara has argued more broadly that a key reason for the lackluster development performance of that country, particularly in the first decades of its independence, has been systematic official discrimination against its ethnic Chinese population, most of whom, until as late as 1975, did not have Filipino citizenship (Yoshihara, 1994: 235). Alasdair Bowie generalized further by arguing (by way of extrapolation from Malaysian evidence) that wherever 'society is communal and [...] there are disparities in economic and political power between communities', tensions among those communities 'are likely to constrain the state's ability to pursue industrial strategies of the kind employed by the capitalist developmental states of East Asia' (Bowie, 1991: 159).
The idea that ethnic capitalism as such -or more exactly, the political reaction against it -has been a central obstacle to development in Southeast Asia is not universally popular. Among its critics is Malaysian political economist K.S. Jomo, who stresses (citing in particular the Christian minority in Korea) that the Northeast Asian countries are themselves 'hardly culturally homogeneous' and suggests that the real key to the Northeast/Southeast Asia divergence lies in the contrasting ways in which their entrepreneurial groups, regardless of ethnicity, were treated in colonial times by their Japanese and European rulers respectively (Jomo, 2003: 18). 'A crucial question is whether or not there is or has been a capable and strong enough national bourgeoisies in the Southeast Asian contexteven if only mainly from among the ranks of the Chinese businessmen -to have been able to effectively advance late industrialization. There are several reasons why the existence of such potential is doubted, beginning with the consequences of European colonialism. Unlike Japanese colonialism in South Korea and Taiwan, European colonial policies in Southeast Asia are believed to have strengthened the development of comprador or dependent elements of the primarily ethnic Chinese bourgeoisie. Hence, such interests have been integrated earlier on [...] with foreign capital, at the expense of other business interests who might have been more inclined to undertake or support nationalist economic projects.' (Jomo, 2003: 19.) Is such scepticism about the industrial potential of postcolonial Southeast Asian capitalism justified? There is little doubt that Japanese colonial policy, especially in Korea, was in various ways more favorable to the industrial sector than were the policies of the European colonial powers in Southeast Asia, and that considerable industrialization did take place in prewar Korea (Haggard, Kang & Moon, 1997: 868-869). But whether the entrepreneurs who actually spearheaded South Korea's postwar industrialization owed anything to this earlier history, or differed markedly in their origins or capabilities from the 'comprador' capitalists of Southeast Asia, is much less clear. Studwell, in a fascinating series of biographical cameos, points to striking parallels between the early careers of prominent Malaysian 'godfathers' and that of Hyundai founder Chung Ju-Yung (1915-2001, whose first business activities were in rice trading and car repair, followed by construction contracts for the US military during the American occupation (1945)(1946)(1947)(1948).
'Chung's family business structure, the nature of his business activities and his foibles were the same as those of the oligarchs who have dominated Southeast Asian economies. In the 1950s in particular he operated in a cronyistic environment in which kickbacks and bid rigging [...] were endemic. By the time of Park Chung Hee's coup, Hyundai Construction was one of the Big Five Korean building firms. Chung had never made or exported anything. He was just a politically astute entrepreneur with a reputation for getting construction jobs done. In post-independence South-East Asia, he would have carried on doing construction work and added more domestic business concessions to his portfolio as the state offered them up.' (Studwell, 2013: 90.) What changed, according to Studwell, was simply that the government of South Korea, unlike those of Southeast Asian countries, decided to induce, indeed force, its 'cronies' to embark on the difficult and risky path of learning and investment which eventually made Hyundai a global manufacturing giant. This view is supported by the insider testimony of a former manager of the (Sino-)Indonesian automotive company Astra, a Toyota import and assembly partner, speaking in 2007 about the differences between his own firm's development and that of Hyundai: 'A top Astra official recounted a meeting in South Korea at which the head of Hyundai Motors emphasized that his company was given no choice but to develop the internal capacities necessary for exporting automotive products with high levels of local value added. By contrast [...] Indonesia's clientelist politics, its large domestic market, and its extensive protection [...] allowed Astra lots of "freedom" to avoid exports while depriving the firm of the support required to improve its internal technological capacities.' (Doner, 2009: 274.) The difference, in other words, lay in the policies of the respective states, and not in the 'strength' or 'capability' of the enterprises or entrepreneurial classes concerned. And in Indonesia and (especially) Malaysia, as we have seen, an important influence on industrial policy was the prevailing system of ethnic business specialization, which made it difficult for politicians to engage established capitalists as partners in national development, rather than as partners in corrupt clientelistic exchanges of licenses for bribes.
In Malaysia and Indonesia the developmental problem sketched above is particularly persistent and intractable given that Islam, since the 1970s a growing influence in society and politics, divides the majority of the indigenous population from the ethnic Chinese. In the case of Buddhist Thailand, where the ethnic boundary is more porous and more people of Chinese descent have assimilated to the majority group, the situation is more complex. Here the moderating effect of cultural affinities was further enhanced from the 1930s to the 1970s by state policies of compulsory cultural assimilation for Sino-Thais, and by growing acceptance among indigenous Thais of the moral legitimacy of business as a profession. Yoshihara already regarded the 'Chinese question' as 'a dead issue' for Thailand in the 1980s (Yoshihara, 1988: 128). Certainly it does not seem to have played a significant role in the country's recent political turbulence, even though the populist leader who changed the face of Thai politics with his election as prime minister in 2001, Thaksin Shinawatra, is himself a Sino-Thai businessman. It follows that the phenomenon of ethnic business specialization provides no direct explanation for the continuing absence in today's Thailand of a concerted national industrialization drive based on the incubation of indigenous design and export capacity.
Indirectly, nevertheless, the dependent pattern of industrial development that is so clearly evident in contemporary Thailand may still reflect a history of ethnic division. Although Thailand's capitalist class is no longer seen today as an alien Chinese intrusion into a society of Thai peasants, aristocrats, and clerks, until the middle of the twentieth century this was emphatically how Thais viewed it, even in official circles (Landon, 1940). It was not until 1957 that attempts to promote government corporations to counter Chinese economic power were abandoned, and residual discrimination in the economic sphere continued until the 1970s (Yoshihara, 1994: 35-36). It is important to remember that because of the selfperpetuating aspect of dependent industrial development as illustrated above, the forces pushing a nation down the dependent path do not necessarily need to be sustained for long in order to make that path inescapable. Once local industries are outcompeted in the domestic market by foreign rivals, or inserted into international production chains under the auspices of foreign partners, a window of opportunity for autonomous development may be more or less permanently closed. To what extent this is what happened in post-war Thailand is a question deserving of further research.

Conclusion
Using selective comparisons with Northeast Asia, particularly Korea, this essay has sketched the outlines of two linked arguments regarding the long-term economic development trajectory of Southeast Asia. First: the persistent inability of Southeast Asian countries to transcend middle income status results partly from their dependence on, and domination by, foreign investment and foreign technology in the industrial export sector. Second: the policy decisions which resulted in this dependency -or perhaps better, which failed to avert it -were related to two major ways in which the historical circumstances of Southeast Asia differed from that of Northeast Asia: the fact that its governments were not under existential political and military pressure to bring about industrial transformation, and the fact that its economies were dominated by alien business minorities.
It should go without saying that by endorsing the significance of ethnic capitalism in relation to the middle income trap, I do not mean to endorse the very practice of 'blaming the Chinese' for Southeast Asian problems which has been so damaging to Southeast Asian attempts to overcome those problems. My point, and that of many other commentators, is simply that the historical concentration of business wealth and expertise in the hands of the Southeast Asian Chinese has made it politically and psychologically difficult for Southeast Asian leaders to emulate their South Korean counterparts by embracing their domestic capitalists, even with the kind of 'tough love' which Park Chung Hee showed for Chung Ju-Yung and his peers, as partners in national development.
The on-going abatement of tensions between indigenous and ethnic Chinese groups in some Southeast Asian countries, particularly Thailand and the Philippines, must bode well for a solution to these problems. However, it is not my purpose here to make predictions on the basis of the arguments presented. Given the theme of the current special issue, it is more appropriate to conclude by reflecting on the relevance of my narrative to the study of economic history of Indonesia during the colonial era.
That narrative, of course, has not been directly concerned with colonialism, and unlike some writers on Northeast and Southeast Asian development performance (Booth, 2007;Kohli, 1994) I have not tried to trace the lineages of key postcolonial policies back to colonial precursors and preconditions. In relation to South Korea I have argued specifically against the idea that post-war industrial success built on entrepreneurial foundations that were decisively stronger than those present in Southeast Asia at the time of decolonization. The military and revolutionary threats that galvanized industrial policy in Korea and Taiwan certainly had little to do with colonial legacies.
When it comes to ethnic capitalism and its consequences, on the other hand, colonialism is clearly implicated in the story. The predominance of ethnic Chinese business in Southeast Asia, after all, was established during the colonial era, and partly as a result of colonial policies. At the genesis of European colonialism in Indonesia during the seventeenth century, the Dutch East India Company fought to monopolize the archipelago's international commerce at the expense of independent trading cities like Makassar and Banten, while populating its own strongholds at Jakarta and elsewhere with Chinese immigrant laborers and shopkeepers (Reid, 1993: 267-325). Later colonial states, particularly in British Malaya, consolidated the pattern by promoting both Chinese immigration and ethnic segregation. It was partly due to such colonial policies that almost every Southeast Asian nation effectively entered the modern world as what Reid, writing of Indonesia, called 'a people without a bourgeoisie' (Reid, 1974: 6). Much more could of course be said about the political and ideological consequences of this circumstance, which often pushed Southeast Asian nationalists in the direction of comprehensively anti-capitalist ideas as well as making it difficult for them to engage with ethnic Chinese capitalism in particular.
When assessing the role of European rule in this story, however, it is important to note that the presence of foreign trading minorities in Southeast Asia antedated colonialism, and to remember that the characteristic modern pattern of ethnic Chinese commercial dominance developed in independent Thailand as well as in the European colonies. Aspects of the region's precolonial political economy already seem to have been conducive to it (Reid, 1993: 124-129). There is in fact good reason to think of this as the global norm. Historically speaking, trade-specialized ethnic minorities have been common features of underdeveloped and developing economies across most of the world -including, until quite recently, Europe, with its prominent, and despised, Jewish commercial diasporas (Hamilton, 1978).
Rather than looking for ways to interpret Southeast Asia's middle income trap as a legacy of colonialism, it may be more productive in this context to ask whether the processes giving rise to the trap had parallels during the colonial period which can shed light on the economic history of that time. Decidedly thought-provoking here is the ease with which foreign capital, in postcolonial Southeast Asia, has apparently been able to dominate key industrial sectors and subordinate domestic capitalists to its own projects and interests, to the extent of preventing them from developing skills and capacities which might enable them to compete with their overseas 'partners'. In colonial times, of course, there were as yet no internationally integrated industrial production chains. But if dependency has been a tangible problem for sovereign nations in the postcolonial period, how much more of a problem is it likely to have been for a colonized people, ruled by a state acting in even greater concert with foreign investors, and in an era when the superiority of foreign over indigenous capital, business methods, and technology was even greater than today?
The usefulness of the Korean counterfactual when it comes to understanding postcolonial Southeast Asia is a reminder that when studying the impact of foreign investment in colonial Indonesia, it cannot be enough simply to note how many jobs, 'linkages' and 'spillovers' were created, or what skills were apparently transferred. We must also try to consider to what extent indigenous competitors, both actual and potential, were crowded out by the presence of foreign capital, and with what long-term consequences. The question is never just what happened because foreign enterprise and the institutions supporting it were there, but also what might have happened had they not been there.