产业内贸易：当前观点和未解决问题【外文翻译】.doc

1、 外文翻译 原文 Intra-Industry Trade: Current Perspectives and Unresolved Issues MaterialSource:http:/ Author: David Greenaway and Chris Milner Contents: I. Introduction. II. Current Perspectives. III. Unresolved Issues. I. Introduction The earliest references to trade between industries appear to be Hilge

2、rdt 1935 and Ohlin 1933. Despite such early recognition, the post-war concern of trade theorists with the refinement of the Heckscher-Ohlin-Samuelson (H-O-S) model, together with the absence of any documentary evidence of intra-industry trade combined to ensure that these early references to the phe

3、nomenon lay dormant for very many years. It is in fact only over the last twenty years in general, and the last ten years in particular that the subject has been intensively scrutinized. The result of this blossoming of interest is that intra-industry trade has been and continues to be a frontier ar

4、ea for international economists. This paper aims to take stock of what is now a voluminous literature in order to comment upon what has been achieved by researchers working in the field and to identify issues which remain unresolved and which are likely to provide a focus for further research in the

5、 near future. We therefore intend evaluating work on intra-industry trade as a complete research programmer and to comment on whether the programmer can be expected to continue being progressive (in the sense of Lakatos). The literature on intra-industry trade has developed along various avenues, wi

6、th the strands of the literature often being really quite disparate. In order to provide some structure to this paper we shall evaluate both current perspectives and unresolved issues by reference to three distinct areas: theory, measurement, empirical evidence and policy aspects. In so doing we hop

7、e to focus more precisely on both current perspectives and unresolved issues than would be the case were we to indiscriminately review the literature in its totality. An attempt will be made to draw together the different strands in the final section of the paper where we “take stock“. The paper is

8、organized as follows. Section II provides a brief historical perspective for the analysis of intra-industry trade. Section III examines current perspectives on theory, measurement, empirical evidence and policy aspects. Section IV identifies unresolved issues in each of these areas. II. Current Pers

9、pectives 1. on Theories of Intra-Industry Trade Considerable progress has been made in modeling intra-industry trade to take us beyond the primitive (in the literal rather than pejorative sense) models of Gray 1973 and Grubel, Lloyd 1975. Intra-industry trade is invariably associated with imperfectl

10、y competitive product markets where consumers have diverse preferences, and the production function is subject to increasing returns, and/or markets are segmented. What we have observed therefore is an extension of the analysis of imperfectly competitive product markets in a closed-economy setting w

11、hich was popularized by Robinson and Chamberlin in the 1950s to an open-economy setting. In summarizing “On Theory“ several observations can be made. First, it is clear that the intuition which several early researchers had that scale economies and preference diversity were related to the explanatio

12、n of intra-industry trade has been given a much more rigorous theoretical foundation. Moreover these factors can be embedded in general equilibrium as well as partial equilibrium models. Second the circumstances under which intra-industry trade can emerge would appear to be quite robust with regard

13、both to demand and supply conditions. Alternative forms of preference diversity can be incorporated, as well as alternative specifications of the utility function. On the production side alternative forms of cost function have been specified with the most recent models incorporating sunk costs. Thir

14、d the role of taste overlap and income distribution has been more clearly elaborated. Finally some of the basic models have been developed to allow for multi-product and multi-national firms e.g. Lancaster, 1984; Lyons, 1984. 2. on the Measurement of Intra-Industry Trade Whereas most research effort

15、 on theories of intra-industry trade has been concentrated in the last six or seven years, work on the measurement of the phenomenon has been steadily conducted over the past quarter of a century. During that time several measures of intra-industry trade and inter-industry specialization have been p

16、roposed and refined in various ways. Without doubt the most widely deployed measure of the former is the Grubel-Lloyd index. This is a measure of the proportion of IIT rather than of the absolute amount of IIT. It tends to be used indiscriminately as a measure of both aspects of IIT. The measurement

17、 of intra-industry specialization is less common than the measurement of intra-industry trade and, as a result, no clear “favorite“ index has emerged. The indices recently proposed by Glejser et al. 1979; 1982 have attracted considerable interest, although as Kol, Mennes 1986 point out the index bea

18、rs a marked resemblance to Balassas index of revealed comparative advantage Balassa, 1965. In summarizing “On Measurement“ it can be reasonably claimed that some progress has been made both in exploring the properties of the more popular indices and in resolving some of their shortcomings. The most

19、fundamental measurement problem is the vagaries of categorical aggregation. Even here, ho wever, some modest progress is evident. 3. on the Empirical Analysis of Intra-Industry Trade The literature on empirical analysis can be broadly sub-divided into studies which are primarily of a documentary nat

20、ure and studies which are primarily of an explanatory nature. The former tend to report the results of calculating intra-industry trade at a particular point or points in time for a given country (or countries). The latter attempt to explain observed country or industry differences in intra-industry

21、 trade and, increasingly, taking an econometric approach. There have been a number of attempts to go beyond casual empiricism and test specific hypotheses relating to the growth or pattern of intra-industry trade. Econometric analysis of this issue is, however, fraught with difficulties, both method

22、ological and practical. For example there are genuine problems in testing specific models which may differ only in assumptions. Even when one has specified a model to be tested, many of the variables included in the model are extremely difficult to proxy, product differentiation and scale economies

23、being the obvious examples. Despite these drawbacks a dozen or so econometric analyses which test hypotheses relating to industry characteristics have now been published. In addition there are a handful of studies which focus on country characteristics. These studies differ quite markedly in their c

24、ountry and industry coverage, the time period encompassed, the specification of their models, the proxies used, and even in the manner in which they measure intra-industry trade. Given such model diversity it would not be terribly surprising if little by way of generalization of results could be mad

25、e. In fact, however, some consistency of results is evident. Taste similarity/ taste overlap, product differentiation, decreasing costs and market concentration appear to be deterministically related to intra-industry trade. There is enough support for these factors, which together with less conclus

26、ive support for other factors, can lead us to reject the view that recorded intra-industry trade is little more than the outcome of random measurement problems. . Unresolved Issues 1. on Theories of Intra-Industry Trade Where theoretical analysis of IIT is concerned it is possible to identify severa

27、l areas where further research is needed. Firstly, as we saw in section III. I. above, research so far has tended to concentrate on extreme cases of market structure. Either markets are assumed to be segmented or fully integrated; entry is assumed to be either completely free or completely blocked.

28、Further analysis of the factors which sustain these particular characteristics is required, as well as the exploration of intermediate cases. Secondly, almost all of the work so far has focused upon single-product firms (with a few exceptions like Lancaster 1984 and Lyons 1984). Thirdly further anal

29、ysis of the factors which support multi-product firms is necessary, if only because there are likely to be asymmetries between incumbent firms and new entrants in such circumstances. Likewise, the existence of multinational firms may be a source of asymmetry between incumbents and entrants and the f

30、actors resulting in multinational involvement in IIT is therefore a third area for further exploration. Fourthly the modeling of conjectural variation on price (as well as other strategic variables such as product specification) could benefit from further refinement. So far only the simplest assumpt

31、ions have been invoked. It is possible that recent developments in the application of game theory could yield benefits here. Fifthly, it is becoming increasingly obvious that a significant proportion of measured IIT is accounted for by trade in parts and components; Most of the models developed so f

32、ar assume trade in final goods. The modeling of trade in intermediates needs to be explored further. Finally the response of producers to alternative sources of supply and the manner in which this is affected by risk, and attitudes to risk could be a fruitful area of research. Progress on these issu

33、es should be possible in the short run if only because researchers have a reservoir of knowledge upon which they can draw. Many of these unresolved issues have been more fully explored in the theory of the firm/industrial organization literature. 2. On the Measurement of Intra-Industry Trade Without

34、 doubt the greatest challenges regarding measurement relate to the problem of categorical aggregation. Despite the progress made on measurement some commentators remain of the opinion that categorical aggregation is the single most important source of intra-industry trade. This extreme position does

35、 not seem to be adopted by the majority of students of the subject. Nonetheless there is a need to learn more about the influence of the phenomenon. Several possibilities exist. Further experimentation with regrouping would be worthwhile. Clearly this option would require a greater degree of consens

36、us regarding “industry“ delineation than exists at present. Also, further experimentation with the measure proposed by Greenaway, Milner 1983 would be worthwhile. Thirdly, factor ratio analysis of the type used by Finger 1975, Rayment 1976, and more recently by Lundberg, Hansson 1986 could pay divid

37、ends. Finally the “residual method“ proposed by Greenaway 1984 has yet to be applied empirically. This too could be worthwhile exploring. Identification of the nature and influence of categorical aggregation is an issue which is fundamental to the entire subject of intra-industry trade. Ironically t

38、he amount of resources devoted to examination of the question is probably a fraction of that devoted to theorizing on sources and policy aspects. This is easily explained by the longer pay-off period for results. It is, however, unfortunate and it is to be hoped that the issue will be the subject of

39、 greater academic scrutiny in the future. 3. On the Empirical Analysis of Intra-Industry Trade Clearly the main thrust of empirical work should be directed at examining industry specific rather than country specific explanatory factors. This being so there are several ways in which progress could be

40、 made. Firstly, in cross section econometric analyses greater attention could be paid to the details of the hypotheses being tested. Frequently a given set of results is consistent with several (sometimes competing) hypotheses. More precise hypothesis testing, and an attempt to discriminate between

41、competing hypotheses should facilitate less ambiguity in interpreting results, as should improved measurement or praying of right hand side variables. A second area of econometric research which is worthwhile is the refinement of simultaneous equation models. Arguably some market structure variables

42、 and IIT should be viewed as endogenous variables and modeled accordingly. Thirdly, with the notable exception of Messerlin and Becuwe 1986 little has as yet been done by way of time series analysis. The way in which patterns of IIT evolve and the forces which shape trends in trade could be profitab

43、ly analyzed. Finally, very little indeed has been done of a “case study“ nature. Industry specific studies are potentially the richest source of information on IIT yet very few industry specific studies exist. At this level of desegregation it is possible to identify quite sharply different product

44、groups and product lines, and to chart changes in product specialization. Moreover, it is possible to identify the role of both industry specific and exogenous variables, and to provide scope for examining the changing nature of industry boundaries. 译文 产业内贸易：当前观点和未解决问题 资料来源 : http:/ 作者： David Greena

45、way and Chris Milner 内容： 一、导言 二、当前观点 三、未解决问题 1 导言 最早提及到产业内贸易的是 Hilgerdt（ 1935） 和 Ohlin（ 1933）。尽管有早期的提及，但战后，贸易理论家关心的是（ H-O-S）模型，结合其他所有产业内贸易的记录，来确保几年前的现象作为证据。它实际上只是在过去的二十年多年，特别是最近的十年 ，这个课题被深入审议，产业内贸易逐渐成为国际经济前沿领域。 本文的目的是通过大量的文献资料来论证在该领域工作的研究人员取得的成就，找出那些现在仍然没有解决的问题，在不久的将来在该研究领域提供进一步的证据。因此，我们打算通过一个完整的研究程

46、序来论证产业内贸易，评论该程序是否还有进步的可能。 对产业内贸易的文学著作已经沿着多种途径发展，而且文学著作水平的高低各不相同。为了使本文结构合理化，我们将对当今问题和未解决的问题的三个不同的方面分别来论述：理论，测量方法，经验和政策。在这样做的同时，我们希望把重点放在当今 观点和未解决的问题上，从总体上对文学著作进行更深一步的理解。尝试在本文最后一节汇集不同程度水平的研究。 本文的结构如下 ： 第二节 为 产业内部贸易分析提供了 一个 简 明的 历史观点 ，第三节审查理论，计量，实证和政策方面目前的观点 ， 第四节确定了这些领域中悬而未决的问题。 2 当前观点 2.1 论产业内贸易理论 产业

47、内贸易的建模发展已经取得了相当大的进展，超越了我们最原始的Gray 1973和 Grubel, Lloyd 1975模型。产业内贸易理论总是和不完全竞争的产品市场 ，消费者偏好联系在一起。产品的功能越来越受到关注，占领了越来越多的市场。因此，我们所观察到的是一个在封闭的经济环境下的不完全竞争的产品市场分析，直到 20 世纪 50 年代，由 Robinson 和 Chamberlin 将其推广到一个开放的经济环境当中。 有几点意见可以在“理论”的总结上得出：首先，很显然，几个 西方 研究者对经济规模和偏好多样性进行了相关的产业内贸易的解释赋予了更为严格的理论基础。此外，这些因素可以融入一般均衡和

48、局部均衡模型。其次在任何情况下，产业内贸易可以出现比较强劲的相当于需求和供给的情况。偏好多样性的 两者可以合并，就比如其他规格的实用功能。在生产方面成本函数的替代已经被最近的成本模型所指定。再次，重叠和收入分配的作用已经得到更清楚的阐述。最后，一些最基本的模型已经发展到适合多种产品和多国公司。 2.2 论产业内贸易的测量 虽然大多数的产业内贸易理论的研究工作集中在过去的六、七年。对这种现象的测量工作一直在过去的四分之一的世纪中稳步进行。在这段时间，多种测量产业内贸易和产业内专业化化的方法已经用多种方式证明和推定。毫无疑问，前者的应用的最广泛的是 Grubel-Lloyd 指数。这是对产业内贸易

49、的比例衡量，而不是绝对衡量。它往往被用来测量产业内贸易的两个方面。产业内专业化的测量比产业内贸易不常见。因此，没有明确的一个最准确的测量指数。有几点合理的解释可以在“测量”上得到总结：在探索一种更受欢迎的指数和解决他们的缺点上都得到了进一步的发展。最根本的测量问题是绝对测量的无法确定性。即使在这里，一些进步都是显而易见的。 2.3 论产业内贸易的实证研究 文学上的实证研究大致可以划分为记录性的研究和解释性研究。前者倾向于计算产业内贸易对某一国家在某一个特定时间的情况。后者试图解释产业内贸易的行业差异，并越来越多的采取经济手段。已经有了大量的尝试企图超越原有经验，测试有关增长或者是产业内贸易的模型。实证研究的重点在于：需要有方式和实践。例如在测试时需要特定问题特殊考虑。即使在一定已制订的特殊模型中，多个变量在模型中试很难证明的，产品差异和规模经济是明显的例子。