1、 外文翻译 原文 High-Tech Exports from Developing Countries: A Symptom of Technology Spurts or Statistical Illusion? Material Source: EBSCO Author: Martin Srholec Centre for Technology, Innovation and Culture (TIK), University of Oslo It is we ll es tablis he d in the literature that ec onomic de ve lopme
2、nt re quires structural change, so it is important to analyse structural shifts along the way. It is also rightly argued that moving into manufacturing exports entails greater deve lopme nt opportunities because of economies of scale and sc ope, e xport diversication and typically higher skills and
3、technological inte ns ity, c ompare d with exports based on exploitation of natural e ndow ments. Although a broad distinc tion betwee n primary, res ource-based and ma nufacturing trade is rela tively stra ightforward, it is far more c omplicate d to differentiate technological intensity of manufac
4、turing trade. As noted above, it is customary to capture technological intensity of exports by comparing specialization in products perceived as high tech across countries. Figure 1 provides a comparison for all countries for which the relevant data are available (a sample of 108 countries in 2003).
5、 The share of high-tech products in merchandise exports is plotted on the vertical axis a gainst s pecialization in e xports of elec tronics on the horizontal a xis. The dotted lines show sample averages, which divide the gure into four qua dra nts with below/a bove a vera ge sc ore s along the two
6、dime nsions to indicate countries specialized in exports of these products. The data for high-tech exports are based on the World Developmen tIndicators data set (World Ba nk 2006), whic h acc ording to the s ource refers to products with high R ra dio, tele vision a nd commu nications equipme nt (7
7、6), elec trical mac hine ry (77) and me dica l, prec ision and optical instruments (87, 881, 884 and 885)all codes according to SITC, rev. 3. A majority of countries maintain a below-average share of high-tech products in exports. Most low-income countries report negligible high-tech e xports, altho
8、ugh spec ializa tion in high-tech e xports is by no mea ns a privilege of the most advanced countries. Striking is the fact that the country specializing the most in high-tech exports is the Philippines, where almost two-thirds of exports fa ll int o the high-tech category. Other outliers inc lude M
9、alta, Singa pore, Mala ys ia a nd Taiwa n, w he re high-tec h products. account for more than a third of exports. A handful of typical examples of developing (or emerging) countries, such as China, Thailand, CostamRica, Me xic o, Hungary a nd Korea als o pe rform quite we ll in the high -tec h m are
10、a. A closer look at the structure reveals that electronics accounts for the bulk of high-tech exports in most countries. The match is particularly strong for the most specialized countries, where high-tech exports are typically dominated by a limited range of electronics products. Overall correlatio
11、n between the s pecialization in exports of high-tech products and electronics is extremely high. Figure 1 shows that these indicators share more than 90 per cent of cross -country variance. Therefore, in the following sections, we focus narrowly on the analysis of trade in electronics as a broa der
12、 presentation of high-tech exports. What does it mean? What does it tell us? Looking solely at the specialization in high-tec h (or electronics) e xports, one c ould easily conc lude that a number of de veloping c ountries ha ve bee n e xtre me ly s uccessful in technologica l catching up. If taken
13、literally, the gure might be interpreted to suggest that these countries have even overtaken the United States, Japanand the European Union in terms of the technological intensity of their economies. Srholec (2006) direc tly compa res R&D inte nsity of the high-tec h sectors in the OECD area. It is
14、shown that some countries with a relatively low income compared with the advanced OECD members, such as Mexico, Poland, Slovakia, Hungary and the Czech Republic, maintain vastly lower R&D intensity in high-tech electronics compared with the threshold that draws a cut-off point in the taxonomy betwee
15、n high tech and the rest of manufacturing (around 20 per cent of R&D in relation to value added according to OECD 2003: 156). It makes it truly problematic to use any high-tech labels in the context of these countries. Although R&D data for de veloping c ountries are sca rce, a limited com paris ona
16、 t leas t at the a ggre ga te le vel ca n be ma de als o for non-OE CD me mbers. Specializa tion in high-tech e xports doe s not matc h the inte nsity of the economy on R&D e xpe nditure in ma ny area s. T he dis parity is mos t striking in the Philippines, Malta, Malaysia, which have more than 45 p
17、ercent of exports in the high-tech elds, and to a lesser extent in Costa Rica, Thailand and Mexico, also with signicant high-tech exports, but in all of these countries spending on R&D remains well below 1 per cent of GDP. He nce the funda me nta ls of s pecialization patte rns of these c ountries a
18、re probably not based on sophisticated technological capabilities. This conrms the expectation that production s yste ms in the high-tech elds beca me internationally fragmente d to the exte nt that c ountries ca n e xport la rge a mounts of high-tech produc ts while actually mastering very limited
19、technological capabilities themselves. The East Asian region clearly emerges as an important cluster for the ma nufacturing of high-tec h products, but there seems to be a fairly s trong divide in localization of related tec hnological capabilities in the “rst tier ”of the ne wly indus trialize d co
20、untries and the re st of the re gion. A group of high-income Asia n c ountries, na mely Singa pore, Taiwa n, Korea and Ja pan, is among the top 15 countries in the world in both specialization in high-tech exports and R&D intensity, while the others fall well behind in technological capabilities. As
21、 a consequence of its sheer size and unique development trajec tory, China is argua bly a special case in this conte xt. In terms of R&D intensity, China has already overta ke n s ome of the high-income countries such as Ireland and some southern European countries. Regional differences between grav
22、itational centres of business activity and other mainly rural areas s ugges t that s ome Chinese re gions proba bly ma intain R&D inte nsity substantially above the national average and even closer to that of developed countries. As suggested, the phenomenon behind the contrast between specializatio
23、n in high-tec h e xports a nd indige nous technology ca pabilities is the increasing fragmentation of value chains, particularly the separation of technological deve lopme nt from rela ted manufacturing activities. T he e m pirical evidence seems to conrm the fact that technologically intensive acti
24、vities are sticky, highly concentrated in space and remain localized in the home areas of large multinational corporations.8 Foreign direct investment in R&D is increasing worldwide, but it is highly concentrated among developed countries (Le Bas and Sierra 2002). Analyses of rm level data even sugg
25、est that foreign afliates are less likely to venture into R&D activity compared to domestic owned rms in countries behind the technology frontier, such as the Czech Republic (Srholec 2005).It is therefore not surprising that R&D intensity remains rather low in ma ny deve loping c ountries des pite t
26、he surge in high-tec h exports, as they attract mainly manufacturing-based fragments of global production systems, while the technology-intens ive fra gme nts re main concentrated elsewhere. 4 The Effect of Fragmentation on Trade in Electronics How could a developing country turn out to be a top exp
27、orter of electronics despite having very limited technological capabilities? Could it be one of the peculiar outc omes of high ope nness to tra de, foreign direct investment and globalization of production in general? In order to throw some new light upon these questions, we look more closely at the
28、 role of the deepening fragmentation by e xa mining intra -produc t tra de in e lectronics. T he dispersion of the various fragments of production systems mirrors in ows of inputs production factors and interme diate inputs suc h as parts and components) betwee n countries, whic h makes it poss ible
29、 to trace fra gmentation in tra de statistics.Following Kaminski and Ng (2001), Yeats (2001) and Lall et al. (2004), we es tablis h a distinction between trade in nal electronics products and tra de in their components. Instead of using their rather arbitrary classication of products in these two gr
30、oups, howe ver, we use the Broa d Ec onomic Categories (BEC, rev. 3) classication ofcially dened by the United Nations, which separates trade in nal products fromnterme dia te inputs for the purpose of national accounts.9 Given the availability of relevant data, the ana lysis provides e vide nce f o
31、r a sa mple of 111 c ountries in 2003. The data are based on the UN Comtrade Database (United Nations 2005) with the only exception being Taiwa n, for whic h fully c ompara ble data are a vaila ble from the OECD International Trade by Commodity Statistics. It is essential to realize that trade stati
32、stics do not measure domestic value added embodied in national exports, but only gross value (sales or turnover) from exports, which allows for a direct link between the imports and exports of a country. The basic idea is that electronics componentsas intermediate inputs ca n be importe d for the as
33、se mbly of nal products or for further processing a nd re-exporte d to serve foreign markets. If components are imported, their full value might be mirrored in export “sales” of a country irrespective of (a possibly low) value actually having been added in the country. The intra-product trade has tw
34、o consequences: (i) the value of world trade is inated by components repeatedly crossing national borders, of which the full value is counted multiple times in trade along the route, and (ii) a direc t link be twee n the value (a nd struc ture) of na tional imports and exports is created with countr
35、ies involved in the intra-product trade appearing to be specialized in certain products even if value a dded domestically to the exports is fairly low. As the de gree offra gme ntation a nd he nce the ination of trade statistics is unevenly distributed across industries a nd c ountries, this effect
36、biases the structure of world tra de and comparis ons of national specialization. 译文 发展中国家的高技术产品出口： 技术喷出的象征还是数据幻想 资料来源 : EBSCO 外文数据库 作者： Martin Srholec 在一些文献中都普遍的认为经济的发展需要结构的改变，所以从事分析结构转变是十分重要的。与依赖于开采自然资源的出口而言，将结构转变移入制造业出口会有更大的发展机会，因为它们的规模经济，出口的多样化和有代表性的较高技能以及技术强度。尽管初级产品，资源依赖型和制造业贸易之间的较大区别是很明显 的，但是如
37、何区别制造业中的技术强度是相当复杂的。 就如上面所说的，通常通过比较产品作为高新技术产品出口到各个国家的专业化来分析出口产品的技术强度，图表 1 提供的是所有国家可以使用的数据的一个比较（ 2003 年 108 个国家的一个样本）。竖轴表示的是高技术产品在商品出口中所占的份额，横轴表示的是电子产品出口是专门化。虚线表示的是样本的平均数，同时也把整个分为四个象限，数据都在平均数的或上或下，两条线反映的是各国在出口这些产品时的专业化。 高新技术产品出口的这些数据是来自于世界发展指标 的数据（世界银行2006），这 是根据与具有高研发投入强度的产品的相关资源而得来的，比如在航空航天领域，电脑，药物，
38、科学仪器和机电产品。这是 OECD 分类学的定义而且相当接近帕维特分类学中基于科学的工业品出口。电子产品的出口数据来自于 comtrade 数据库（联合国 2005），并且覆盖了接下来几种产品的贸易领域：办公室，会计和计算器械（ 75）；收音机，电视和通讯设备（ 76） ，电子机械（ 77）和医药，精密光学仪器（ 87， 881， 884 和 885）， 所有的编码根据 SITC 第三条。 大部分国家的高技术产品出口所占的份额都是低于平均水平的，多数低收入国 家显示微不足道的高技术产品出口份额，尽管高技术产品出口中的专业化绝不是发达国家的特权，但是事实很明显国家专门化大部分投入在高新技术产品出
39、口的是菲律宾，菲律宾国家的三分之二的出口是高新技术出口。其他局外人包括马耳他，新加坡，马来西亚和台湾，他们的高技术产品出口超过三分之一。少数典型的发展中国家，比如中国， 泰国， 哥斯达黎加，莫斯科，匈牙利和韩国在高新技术领域都发展的挺好的。 更近一步的观察这个结构，发现大部分国家的电子产品在高新技术产品中占了很大比例。对于许多专业化的国家来讲这个比较是相当强烈的。他们都是通过有限制的电子贸易来取得占有主导地位的优势。高新技术产品的出口专业化与电子贸易的出口专业 化 之间的全部联系是极其紧密的。图表 1 中反映的是这些指示器占到全国城市变化的 90%。因此，在接下来的调查研究中，我们比较集中于将
40、电子贸易作为高新技术产品出口的典型代表来进行分析研究。 图表 1， 2003 年电子产品出口与高新技术产品出口的专业化 低收入 中收入 高收入 高收入大于 15000，中收入在 5000 和 15000 之间，低收入小于 5000 数据来源：世界银行（ 2006）和美国（ 2005） 这意味着什么？她要告诉我们什么？单独的来观察高新技术产品（或电子贸易）出口的专业化，能够很简单的总结出一点，一定数量的发展中国家在技术更新方面已经相当成功了。如果逐个来看，图表可以解释为这些国家在经济技术强度方面已经赶上了美国，日本以及欧盟。 Srholec(2006)直接将 OECD 领域内的高新技术研发投入进
41、行比较。结果显示一些相对低收入的国家与那些 OECD 中发达的国家比如莫斯科，波兰，斯洛伐克，匈牙利和捷克共和国相比，在高新技术研发投入方面仍然很低， 比较高新技术与其他制造业之间在分类学上有一个临界点（根据经济合作与发展组织OECD 2003： 156，大约 20%的研发投入与增加值有关）。这是有问题的在这些国家背景中使用高新技术的标签。 尽管关于中国家的研发投入的数据还缺乏，在总体水平上说这是一个有限制的比较，但是同样适用于非经济合作与发展组织的成员。在许多领域高新技术产品出口的专业化不能与国家在研发投入方面投入的费用相匹配。在菲律宾，马耳他，马来西亚这些在高新技术领域出口超过 45%的国
42、家差距是很明显的。在哥斯达黎加，泰国和莫斯科等较小的程度上也有显著的高 新技术产品出口，但是所有的这些国家花在研发投入上的费用仍然不到 GDP 的 1%。因此这些国家的专业化模型的基本原理可能不是依赖于尖端的科技能力这个确认了一个期望，高新技术领域 内的生产系统会到达国际性破碎的程度，那样，国家能够出口更大数量的 高新技术产品，而他们自己却实际上控制了非常有限的技术能力。 东亚地区是一个高新技术产品制造商的重要聚集地，但是在新型工业化国家和其他地区的“第一阶梯”中有着相当强壮的本土化技术能力。一些高收入的亚洲国家，换句话说新加坡，台湾，韩国和日本，他们是世界上在高新技术出口专业化和研发投 入强
43、度顶尖的 15 个国家中的几个，而其他国家的技术能力不如这些国家，由于他 们 的规模经济和独特的发展模式，在这个程度上中国是一个特殊的案例，在研发投入强度方面中国已经超过爱尔兰和南欧的一些高收入国家。商业活动中心地区与一些主要的农村地区之间的地区差异使得中国地区可能仍然维持研发投入大体上在全国平均水平之上，甚至可能更接近于发达国家。 正如所建议的，在高新技术出口专业化与本土技术能力比较的背后存在的现象是价值链的多元化，特别是将技术发展与相关制造活动分离开来。经过实证有这么一个事实，技术加强活动是有粘性的，在空间 中高度集中的，而且仍然集中于大型跨国公司的国内市场。在研发方面的外商直接投资是在世
44、界范围内增加的，但是也高度集中与发达国家（ le bas and sierra 2002） .分析公司水平数据显示与走在科技前沿的国内公司相比，外国机构不大可能会在科技研究方面冒风险的。比如像捷克共和国（ Srholec2005）。虽然高新技术产品出口激增，但是研发强度仍然低于许多发展中国家是不足为奇的，因为他们吸引全球生产系统中以制造为基础的分类，然而技术密集型的仍然集中在别处。 3.在电子贸易中分裂的作用 在拥有非常有限的技术能力的 前提下，一个发展中国家如何成为一个顶尖的电子出口商？这是贸易的高度开放性、外商直接投资和全球生产带来的特殊结果吗？为了阐述这些问题，我们通过检查电子贸易的产品
45、内贸易来更清楚的指导深度分裂的作用。生产系统多样化的分裂反映国家间的输入（生产要素和中间媒介的输入），这使得在贸易统计资料中描述分裂成为可能。 跟随 Kaminski和 Ng(2001),Yeats(2001),Lallet al.(2004),我们区分了电子贸易中的最终产品与中间产品，而不是使用这两个群中固定的分类，然而我们使用联合国正式定义的广义的经济 目录分类，它在中间投入中将中间产品贸易分离开来，这是出于国民经济核算的目的。鉴于相关数据的可使用性，这个分析为 2003 年的 111 个国家提供了证据。这些数据来源于联合国贸易数据库（联合国 2005），台湾是一个例外，来自 OECD 国
46、际商品贸易统计是可比较的数据是有用的。 事实上应该明白贸易统计的数据不能反映在国际贸易中本国的增加值，但是只有出口贸易总值（销售或者营业额）能够在一个国家的进口和出口之间建立一种联系。基本的理念是电子零件作为中间媒介能够被做为生产最终产品而进口或者为了进一步的工序和为了外国市场而再进口 。如果进口组件，他们的全部价值可能反映在一个国家无关重要的价值里面，事实上却已经被贴加到一个国家中。 内部生产贸易具有两个影响： 1.是世界贸易的价值被组件在跨越国界的过程中而抬高，他的全部价值被多次计算。 2.国际进出口值之间的直接联系是由涉及到内部生产贸易的国家所创造的，在一些产品上表现出专业化，然而出口的国内增加值是非常低的。因为分裂的深度和今后贸易统计的膨胀将不平衡的对国家和行业产生影响。这使得世界贸易的结构和国家专业化的比较产生了偏见。
