《在工业化国家的高新技术产品出口和经济增长外文翻译（已处理） .doc》由会员分享，可在线阅读，更多相关《在工业化国家的高新技术产品出口和经济增长外文翻译（已处理） .doc（11页珍藏版）》请在三一办公上搜索。
1、在工业化国家的高新技术产品出口和经济增长外文翻译 本科毕业论文外文原文外文题目: High-tech exports and economic growth in industrialized countries出 处: Economics作 者: Martin Falk原文:High-tech exports and economic growth in industrialized countriesMartin FalkAustrian Institute of Economic Research WIFOAbstract The present article provides new
2、 evidence on the impact of the change in the high-tech export share on economic growth in OECD countries. We estimate a dynamic growth model on panel data for 22 OECD countries for 1980?2004, in which the data is measured as 5-year averages. Using the system GMM panel estimator, which corrects for s
3、imultaneity, we find that both business R&D intensity and the share of high-tech exports are significantly positively related to the GDP per working age population. The estimated elasticities are rather sizable but the magnitude suggests that business R&D intensity is more important than the share o
4、f high-tech exports in explaining GDP per working age population.1. Introduction The share of high-technology exports in total manufacturing exports has increased considerably in OECD countries throughout the last 25 years. Some OECD countries such as Finland, Ireland, Korea, the Netherlands and the
5、 United Kingdom have increased their high-tech export share more than other countries. Similarly, the change in the ratio of business expenditures on R&D BERD to GDP has also unevenly evolved across OECD countries. The differences in the evolution of high-technology export share and business R&D int
6、ensity have attracted substantial policy interest. Business R&D intensity is commonly identified as the main input factor in the innovation process, while the share of high-tech exports is regarded as an important measure of innovation output. Both measures are widely used in order to benchmark a co
7、untrys innovation performance. Given the interest in the specific innovation indicators, it is natural to ask as to which of the two factors is more important for economic growth. The related empirical literature agrees that investment in R&D is one of the most significant factors affecting the diff
8、erences in GDP and productivity growth Bassanini and Scarpetta, 2001.However, the relationship between high-tech exports and economic growth in industrialized countries remains somewhat unclear. Crespo-Cuaresma and Wo rz 2005 found that the export share of technology-intensive industries is signific
9、antly positively related to the GDP per capita only for the sample of non-OECD countries, but not for the sample of OECD countries. Using the data for OECD countries, Peneder 2003 finds that exports of technology driven industries have a positive and significant impact on the level and growth of the
10、 GDP per capita. What is common to all in the previous studies is that they suffer from an omitted variable bias since they exclude the R&D intensity as a measure of innovation input. The aim of the present article is to provide new insights into the impact of high-tech exports on economic growth. T
11、he growth equation is estimated using the system generalized method of moments GMM panel estimator based on panel of 22 OECD countries for the period 1980 to 2004, in which the data are measured as 5-year averages.2. Empirical Model The empirical model is based on the human capital augmented Solow m
12、odel that was introduced by Mankiw et al. MRW 1992. Nonneman and Vanhoudt 1996 extended the MRW model by adding the ratio of R&D to GDP. We further augment the MRW model by adding a measure of innovation output such as the high-tech export share as an alternative to R&D intensity. Following Caselli
13、et al. 1996, the steady state level of GDP per capita based on panel data can be described as: where is per capita GDP of the working age population expressed in 1995 purchasing power parities in country i in period t, where is a country specific effect, is a period-specific effect, and it is an err
14、or-term. The set of explanatory variables includes the investment ratio, INVit, the ratio of business enterprise R&D expenditures to GDP BERD, RDGDPit, average years of education in the working age population from 25 to 64 years of age taken from Barro and Lee, 2000, EDUit and the share of high-tech
15、 exports to total manufacturing exports We can derive the regression equation by taking first differences in order to remove unobserved time-invariant, country specific effects for the sake of notational convenience, x shall comprise the explanatory variables: Assuming the residuals of the level equ
16、ation are serially uncorrelated, the values of y lagged two periods or more can be used as instruments in the first-differenced equation. This implies the following moment condition:In order to deal with the potential endogeneity problem, we assume that the explanatory variables in x are predetermin
17、ed, rather than strictly exogenous, implying in turn the following moment conditions:The estimation equation and moment conditions are estimated using the system GMM estimator proposed by Blundell and Bond 1998. This requires following additional level moment conditions:And3. Estimation Results Tabl
18、e 1 shows the estimation results for the growth equation using three different specifications. In order to reduce the influence of potential outliers, we exclude data points whose standardized residual falls outside the interval from _2 to 2. This reduces the sample by seven observations and leaves
19、us with 96 observations.2 In all cases, the Sargan test of over identifying restrictions cannot reject the null hypothesis that the instruments are uncorrelated with the error term at the 5% level.As was expected, we find that the export share of high-technology industries enters the growth equation
20、 with a positive sign and is significant at the 1% level see specification ii. The short and long-term elasticities of GDP per working age population with regard to the share of high-tech exports are 0.025 and 0.29, respectively. The ratio of business R&D expenditures to GDP is also significant at t
21、he 1% implying that increased R&D activities have a significant positive impact on GDP per capita growth see specification i. The short and long-term elasticities of business R&D intensity with regard to GDP are 0.021 and 0.21, respectively. When both variables are included, the coefficient of the s
22、hare of high-tech exports drop from 0.025 to 0.011, and the R&D coefficient drops from 0.021 to 0.014 see specification. Interestingly, the joint effect of both R&D intensity and the high-tech export share is equal to 0.025. However, the point estimate of high-tech exports is more affected than that
23、 of business R&D when each of the variables enters the regression equation. This indicates that business R&D intensity is more important than technological specialization in explaining economic growth. Furthermore, the SE of both of the coefficients is enlarged due to collinearity. Therefore, we als
24、o provide the Wald-test statistics of joint significance indicating that both business R&D intensity and the share of high-tech exports are jointly significant at the 5% level.4. Conclusions The present article provides new evidence of the impact of the share of high-tech exports and business R&D in
25、tensity on economic growth. Failure to control for innovation input can cause high-tech exports to pick up their effect, leading to an overestimation of the impact of the share of high-tech exports. In using the system GMM estimator, which enables us to control for simultaneity bias, we find that bo
26、th measures of innovation performance are powerful determinants of economic growth when entering the growth equation separately. When both indicators are included in the growth equation, we find that business R&D is more important than high-tech exports in explaining economic growth. Therefore, the
27、failure to account for the impact of innovation input leads to an overestimation of the impact of high-tech exports. The results of the present study are important for policy makers, as both business R&D intensity and the share of high-tech exports are one of the main indicators of the European Inno
28、vation Scoreboard EIS, which is published by the European Commission. Given the results, we suggest that business R&D intensity should be given more weight in the composite summary innovation index that is published in the EIS.Appendix译文:在工业化国家的高新技术产品出口和经济增长马丁.福尔克 奥地利经济研究所(WIFO)摘 要 本文为高新技术出口份额的变化对经合
29、组织国家的经济增长产生影响提供了新的证据。我们根据22个经合组织国家1980-2004年的数据估计出一个动态的生长模型,该测量数据为5年的平均数。使用修正性的系统广义距估计方法,我们发现企业R&D强度和高新技术产品出口份额都与工作年龄人口的人均GDP呈现显著正相关。被估计的弹性相当大,但是幅度上显示在影响工作年龄人口的人均GDP上,企业R&D强度比高新技术产品出口份额的因素更重要。引言 在过去的25年里,经合组织国家的高新技术产品出口份额在整个制造业出口份额中大幅增加。一些经合组织国家例如芬兰、爱尔兰、韩国、荷兰和英国已经增加了他们的高科技出口份额并超过其他国家。同样,企业研发支出占GDP的比
30、例的变化在经合组织各国之间也发展的很不平衡。高科技出口份额的发展和企业R&D强度之间的不同吸引了大量的政策兴趣。企业R&D强度一般定义为在创新过程中的主要投入因素,而高新技术产品出口份额被认为是一种重要的衡量创新成果的指标。这两项指标都被广泛使用,以衡量一个国家的创新绩效。鉴于对具体的创新指标的研究,最主要的问题是对于经济增长这两种因素哪种更重要。 相关实证文献认同对R&D的投资对GDP和生产力增长(Bassanini和Scarpetta,2001年)都有不同程度的影响。然而,高新技术出口和经济增长之间的关系在工业化国家还是有些不清楚的。克雷斯波和沃尔斯(Crespo-Cuaresma and
31、 Worz,2005)发现,技术密集型产业的出口份额和人均GDP呈现显著正相关,但是不适用于经合组织国家,而仅仅只适用非经合组织国家。潘纳德(Peneder,2003年)发现使用这些数据用于经合组织国家时,技术出口带动产业对人均GDP水平的增长有着积极而显著的影响。在以前的研究中普遍存在一个被忽略的变量偏差,直到它们排出了R&D强度作为一种创新投入的指标。 本文的目的是提供新的视角研究高新技术产品出口对经济增长的影响。基于22个经合组织国家在1980-2004年之间的面板数据,使用系统广义距估计方法估计出生长方程,该测量数据为5年的平均数。实证模型 实证模型是基于人力资本增加索洛(Solow)
32、模型,而它又是来源于MRW模型(Mankiw-Romeer-Weil model,简称MRW模型)的。南那曼和凡浩特(Nonneman and Vanhoudt,1996)通过加入R&D占GDP的比重这一因素扩展了MRW的模型。我们通过增加一项衡量创新产出的指标,如选择高新技术产品出口份额作为R&D强度,来进一步扩展MRW模型。以下是凯瑟琳等(Caselli,1996)建立的方程,基于面板数据人均GDP水平的稳定状态可以被描述为: 其中是由工作年龄的人均GDP表示的1995年国家i在t期的购买力平价,其中代表一个国家具体的影响,代表一个时期的特定作用,然后是一个残差项。该解释变量包括投资比例,
33、INVit代表企业的R&D支出占GDP的比重(企业研发支出),RDGDPit代表在工作年龄人口(25岁至64岁)中受教育的评价年数。摘自白若和李(Barro and Lee,2000),EDUit和高新技术产品出口占整体制造业出口的份额由表示。 我们可以通过采用第一个差异项消除不可观测且时间不变的,即国家的具体影响来得出回归方程。(为了方便记录,x包括解释性变量): 假设方程的残差连续不相关,滞后的y值在两期或更长时间内可作为在第一个差分方程的工具。这意味着一下矩条件: 为了应对潜在的内生性问题,我们假设解释变量X是已知的,不是严格外生的,这意味着转变一下的矩条件: 使用布伦德尔和庞德(Blu
34、ndell and Bond,1998年)提出的系统广义距估计方法估计方程和距条件。这需要以下其他级的矩条件: 和估计结果 表1显示使用了三种不同的方式得出增长方程的估计结果。为了减少潜在异常值的影响,我们排除了标准化残差落在-2到2区间之外的数据点,这就减少了7种样本结果,然后剩下96种。在所有情况下,确定限制的Sargan测试为不能拒绝与5%水平残差项不相关的零假设。 正如所预料的,我们发现高新技术产品的出口份额积极的进入生长方程,并且有1%水平的显著性相关。工作年龄人口的人均GDP与高新技术产品出口的份额短期和长期的相关弹性值分别为0.025和0.29。 企业R&D支出占GDP的比重也是
35、1%水平显著相关,意味着增加的R&D活动对人均GDP增长具有显著的积极影响。企业R&D关于GDP的短期与长期弹性强度分别为0.021和0.21。当两个变量都计算在内时,高新技术产品出口份额的系数从0.025下降到0.011,而R&D的系数从0.021下降到0.014。有趣的是,R&D强度和高新技术出口份额的联合效应都等于0.025。然而,当每个变量各自进入回归方程时,高新技术产品出口的点估计比企业R&D的影响更大,这表明企业R&D强度在解释经济增长方面比技术专业化更重要。此外,由于共线性双方的SE系数都增大,因此,我们还会提供具有联合意义的沃尔德试验统计方法来表明企业R&D强度和高新技术产品出
36、口份额有5%水平的显著相关联合。 表1结论 本文为高新技术产品出口份额和企业R&D强度对经济增长的力度具有影响提供了新的证据。如果不控制创新投入会导致高新技术产品出口增长迅速,导致高估高新技术产品出口份额的影响。在使用系统广义矩估计方法时,我们能够控制同时性偏差,我们发现,这两项创新指标分别进入生长方程时都对经济增长有巨大的影响作用。当这两个指标同时包括在生长方程里时,我们发现在解释经济增长时,企业R&D强度比高新技术产品出口更重要。因此,未能清楚创新投入的内容时,将会高估高新技术产品出口的影响力。 本文研究的结果对政策制定者有重要的意义,企业R&D强度和高新技术产品出口份额都是欧洲委员会制定的欧洲创新指标(EIS)的主要内容。根据给出的结果,我们建议企业R&D强度在EIS上发表的综合创新指数总结中应该得到更多的重视。 附录 表A1
