|
【Alice Amsden】 The Rise of the Rest: Challenges to the West from Late-Industrializing Economies →【討論】
Implications of the Approach
1. The Role of the State
Professor Amsden began her discussion of the implications of this conceptualization of economic development by noting that if knowledge is perfect, one can make an argument based on Pareto-optimality which largely eliminates the need for state intervention. If two firms have the same productivity but one is in Japan and the other in the Philippines, the one in Japan will be incapable of competing. In contesting this claim, she invited us to think of the labor-intensive cotton textile industry, an industry that has been used by LDCs as a leading sector to generate economic development. During the 1960s, the Philippines, Korea, Taiwan, and Japan were all producing cotton textiles. The Korean and Taiwanese industries (which, paid for by American aid, were fairly modern) were heavily protected, essentially to keep out the ferociously competitive Japanese textile industry. Even in this highly labor-intensive industry, then, Korea and Taiwan were incapable of competing against the intense accumulation of knowledge-based assets and higher productivity of Japan. If information is perfect, she noted, this outcome should not be possible. Further, the classical solution (dating from Ricardo) to the problem facing uncompetitive firms in Korea and Taiwan is to let wages fall (through outflows of money and resultant decreases in prices and wage levels) until competitiveness is restored. The only role for the government is thus getting out of the way and allowing this to happen.
If knowledge is imperfect, however, there is no Pareto-optimality and no way to decide whether the government should try to increase productivity or to decrease real wages. In her theory, then, governments face a real choice. To illustrate this claim, Professor Amsden drew a graph on the board which plotted wage per worker against labor per unit output, and then drew two curves representing Japan's higher-productivity, higher-wage system and the low-productivity, low-wage system prevailing in Korea and Taiwan (xxfigure one). In such a situation, she argued, there is nothing the International Monetary Fund (IMF) or World Bank can tell Korea a priori about whether it should attempt to compete by raising productivity or reducing wages.
It is this real choice between productivity increases and lower wages, Professor Amsden argued, which explains government intervention in late-developing countries. It is this function, rather than reducing externalities or coordinating firm activities (as prevailing arguments in the literature which she finds somewhat ridiculous suggest), that has been crucial. She stressed here again that even in such a labor-intensive industry as cotton textiles Korea and Taiwan were incapable (despite the extreme repression of labor in both countries) of competing with Japan. In order to deal with this phenomenon, an entirely new approach to the role of the government in the economy is required.
2. Why some countries but not others?
If government has been a central part of economic development, an obvious question is why governments succeeded so well in some countries but failed so completely in others. The answer, Professor Amsden suggested, lies in the nature of intervention: governments in successful countries intervened to increase learning and productivity and to make it profitable to invest in knowledge-based assets. They accomplished this through the mechanism of performance standards under the slogan "nothing given for free", and much of Professor Amsden's book takes up the nature of the performance standards that have been imposed. In her talk, she touched upon three kinds:
i. Technostandards. In almost all successful cases, governments provided subsidies to invest in industry and technology, primarily through development bank loans with soft interest rates. Professor Amsden's research in the archives of these development banks indicates that there were huge performance standards attached to these loans, standards which included requirements that firms hire chief financial officers from outside the owning family, that they invest in the latest machinery, that they bring in foreign technical consultants, and so forth. Similarly, governments insisted upon the establishment of formal accounting systems and relations of accountability within firms. The "managerial revolution" was thus brought to late-developing countries through the mechanism of government lending.
ii. Policy standards. Standard economic analyses generally claim that countries such as Korea, Taiwan, and Thailand have succeeded economically because they were open. Professor Amsden argued rather that they were "selectively secluded", noting that "Korea protected anything that they could make money in. So did Taiwan." The Korean automobile industry, for instance, which consisted of just two firms (Daewoo and Hyundai) in a highly protected market, was a recipe for disaster in the eyes of economists. The industry succeeded, however, because of governmental performance standards that insisted that firms must export. Long-term capital loans were provided only on the condition that firms build capacity which could be used partially for export. Professor Amsden added here that because economists do not analyze the connection between exports and long-term investment capital (as the latter cannot be cleanly categorized as promoting import-substitution or exports), the basis of the export drives in these countries does not even appear in economic analyses.
iii. Performance standards relating to firms. Noting the current popularity of analyses focusing on "corporate governance", Professor Amsden argued that this third kind of performance standard was critical in determining systems of corporate governance in late-developing countries. Korea and Taiwan, for instance, have historically had firms with very high debt-to-equity (d/e) ratios, while d/e ratios in Brazil and Mexico have been comparatively low. The difference is explained by the policies development banks in different countries have adopted towards debt-equity ratios. If a government wants to create firms which can compete at the technological frontier against firms from developed countries, Professor Amsden argued, it is necessary for it to encourage high d/e ratios, as internal savings will simply not be sufficient for this project. A policy that promotes low d/e ratios, on the other hand, will lead to an economy dominated by small and/or foreign firms. As a result of these divergent policies, it is now very difficult to think of globally-known "national champion" firms from Brazil or Mexico, while Asia has produced many such firms. Professor Amsden noted the irony that in Brazil, a country with a highly inegalitarian distribution of income, the government deliberately kept d/e ratios low in the interests of equal income distribution.
Summarizing her comments to this point, Professor Amsden stressed that her theory of economic development renders government intervention into the economy much more comprehensible than do competing theories. Governments intervene because firms cannot compete, and will only be able to compete if the government supports productivity increases (though late-developing country governments have also, of course, sought to keep wages low). Again, Professor Amsden stressed that in a world characterized by imperfect knowledge, no argument from Pareto-optimality can be made against this kind of government policy.
3. Country Experiences
Professor Amsden turned next to the question of why some countries were able to enact these policies while others weren't, and to how we might be able to predict (post facto) which countries would be successful in this endeavor. Referring again to North's NIE approach, which explains differential success in terms of different structures of property rights, she argued that these differences are almost impossible to measure and that she doesn't think they can explain the variance in outcomes. In addition to property rights, a country which seeks to develop must possess a legal/social/cultural system supportive of learning, and the skills and management experience necessary for developing knowledge-based assets. Such a system only emerges over time, and Professor Amsden argued that the prior acquisition of manufacturing experience has been critical. Countries which did not possess a substantial manufacturing sector before World War II were highly unlikely to develop one after the War.
Professor Amsden supported this argument by describing the results of a simple regression equation which sought to determine whether the 1950 per capita level of GDP or that of manufacturing is a better predictor of 1994 per capita manufacturing output. The regression, which examined 29 countries (n=29), thus took the value of manufacturing output per capita in 1994 as the dependent variable, and the 1950 per capita levels of manufacturing and of GDP as the independent variables. She noted that even she had been surprised at how well this regression worked, with the adjusted r-squared for GDP per capita turning out to be 0.35 and that for manufacturing per capita to be 0.75. Professor Amsden concluded from this that "manufacturing experience matters", and stated that while organizations like the World Bank focus on "getting the prices right", her approach stresses rather the process of accumulating experience and institutional development.
Manufacturing experience alone, however, is insufficient to explain late development; the type of manufacturing experience has also been important. Professor Amsden mentioned three different forms of such experience. The first, which she described as pre-modern, describes traditions of manufacturing beyond the household and oriented to the market that existed in a number of countries prior to the onset of industrialization -- a "deep source of artisans". The Ottoman Empire, India, and China all had such backgrounds, while Chile, Argentina, Korea and Taiwan did not. The second form, emigre manufacturing experience, she divided into two sub-classes: countries where manufacturing had been established prior to World War II by Europeans or Americans, and those where it had been set up by Chinese. Almost all Latin American countries that accumulated pre-war manufacturing experience fit into this mold. China, meanwhile, was a major teacher in Thailand, Indonesia and Malaysia. The third form of pre-war manufacturing experience was colonial.
The reason these three types mattered, Professor Amsden argued, is that colonialism was very important. Argentina, Chile, Brazil and Mexico all decolonized in the early 19th century, and as a result did not experience the type of build-up of manufacturing experience associated with colonial empires. Further differences exist within the category of colonialism, and particularly in the process of decolonization. The sudden withdrawal of Japan from its empire, in particular, left behind knowledge-based organizations no longer controlled by the Japanese which became the basis for the subsequent development of the private sector in Korea and Taiwan. Japanese domination of the economy thus disappeared at a stroke, a process replicated in the Dutch withdrawal from Indonesia and the Chinese revolution. For all the asset-stripping engaged in by Dutch firms, Professor Amsden claimed, the 400 companies they left behind in Indonesia formed the basis of subsequent Indonesian economic development. The divestment of British companies from India presents a mixed and complex case. India did not expropriate British capital, but that capital was sufficiently frightened of the decolonization process that they tended to divest. However, the central point is that Korea, Taiwan, China and India, which all experienced sudden decolonization after World War II, all developed relatively independent national champions, while the economies of Argentina, Brazil, Chile and Mexico became completely dominated by multinational corporations (MNCs) at a very early stage. Malaysia, Indonesia, and Thailand had more mixed experiences with imperialism and decolonization, and it's more difficult to say what will happen to them in the future, though Professor Amsden suggested that Indonesia's experience with decolonization suggested that it might eventually join Korea and Taiwan.
This analysis obviously depends on the assumption that national ownership of industry is important, a claim which is often contested in this "global" age, and so Professor Amsden touched briefly on the question of whether this is in fact the case. Her major argument that national ownership is crucial was based on the observation that the percentage of research and development undertaken by MNCs outside their host countries remains tiny. The importance of the location of R&D varies by industry: in labor-intensive industries like cotton textiles, for instance, MNCs should be welcomed as bringing jobs and capital. In industries like automobile manufacturing that are characterized by first-mover advantages, however, national ownership is very important, and "nationality matters". Professor Amsden noted in this context that R&D in Latin America is negligible, even "pathetic", and that there is very little development of knowledge-based assets there. Taiwan and Korea, however, have had very different experiences.
Professor Amsden's final point was that income distribution plays a huge role in her book. The breakdown of countries according to decolonization process is replicated in the distribution of income: Korea, China, Taiwan, and India have relatively equal income distributions, while Argentina, Brazil, Chile and Mexico have quite unequal ones. The question then becomes: why do nationally-owned firms and relatively equal income distributions go together?
In conclusion, Professor Amsden noted that her book starts from Alexander Gerschenkron's premises. Like Gerschenkron, she argues that comparing the process of industrialization in, for instance, Mexico and the United States is absurd because the US started so much earlier. Major companies in the United States have all been based upon innovation, the development of new products and knowledge-based assets, while Mexico's experience (like that of Japan, China, India and so on) has been completely different. However, she disagrees with Gerschenkron's famous assertion that the extent of government intervention increases with the lateness of development. This argument, she noted, both implies that contemporary industrialization should require almost total government control of the economy, and underplays the degree of intervention involved in British industrialization (particularly by the navy). Korea has not necessarily been more interventionist than Britain has: it has simply intervened in different ways. Professor Amsden does argue, however, that the later industrialization occurs, the greater will be the degree of foreign control of industry, because of the difficulties of competing against incumbent foreign companies in oligopolistic markets.
＜記録：Derek Hall＞
|
