《服务贸易的开放度与经济增长外文翻译.doc》由会员分享，可在线阅读，更多相关《服务贸易的开放度与经济增长外文翻译.doc（16页珍藏版）》请在taowenge.com淘文阁网|工程机械CAD图纸|机械工程制图|CAD装配图下载|SolidWorks_CaTia_CAD_UG_PROE_设计图分享下载上搜索。
1、效劳贸易的开放度与经济增长外文翻译 外文翻译原文 Openness in Services Trade and Economic Growth abstract This paper examines the relationship between openness in services trade and economic growth. We estimate a threshold regression model to test whether openness in services trade has a different impact on low- and high-in
2、come countries. We consider openness in both telecommunication and financial services. Results confirm the existence of a two- regime split with low-income economies benefiting from greater openness in telecommunication services and high-income economies from financial services openness. Key Words:
3、Services trade, telecommunication services, financial services, economic growth.Openness in Services Trade and Economic Growth1. Introduction The costs and benefits of openness to international trade for economic growth have been studied extensively. Diverse measures of openness to trade have been u
4、sed referring mainly to trade in commodities and include, inter alia, the share of exports/imports in GDP, measures of the prevalence of tariffs/quotas in commodity trade, or general qualitative indicators of openness. Studies, on the whole, demonstrate the positive effects of international trade e.
5、g. Dollar, 1992; Frankel and Romer, 1999; and Sachs and Warner, 1995 though dissenting views have been voiced e.g. Rodrguez and Rodrik 2001. To date, very little has appeared on the growth effects of openness in services trade. This should, perhaps, not be surprising given that services constitute a
6、 new dimension in regional and multilateral trade agreements, especially within the auspices of recent and proposed WTO trade rounds. The view that greater openness in services trade is associated with higher growth seems to have taken hold amongst some policy makers recentlyEmpirical evidence on th
7、is issue, however, has been scant. A recent article in The Economist2004 emphasizes the fact that a well-established infrastructure is necessary in order for openness in services trade to have a positive impact on growth. The article is concernedwhether“ICTsinformation and communication technologies
8、 might have less impact on productivity in poor countries than in rich countries and “ whether a certain threshold level of adoption is required before the productivity benefits of ICTs kick in. It questions whether the benefits from liberalizing trade in services are the same for low- and high-inco
9、me countries. The objective of this paper is twofold: first, to add to the very limited evidence on the effects of liberalizing trade in services on economic growth; second, to test the hypothesis that a threshold level of development per capita income exists in the relationship between services tra
10、de openness and economic growth, and the corollary that the benefits from service trade liberalization do not materialize until such a threshold is achievedWe focus on two service industries: telecommunication and financial services. To test the underlying hypothesis we use the threshold regression
11、TR model, an econometric technique that selects endogenously and tests simultaneously the existence of a threshold level of per capita income.2. The threshold regression model and openness in services trade Hansens 2000 TR model allows for an endogenous sample splitHe develops a sequential search al
12、gorithm to determine the value of the threshold and a Lagrange multiplier test for its significance. Papageorgiou 2002 uses the TR model in the context of a growth regression and finds that initial income per capita is a significant first-round threshold indicator. In subsequent rounds, literacy and
13、 general openness to trade the share of exports plus imports in GDP are also significant threshold indicators. In our model, the threshold variable is initial income per capita and the underlying hypothesis is that the impact of service trade openness on economic growth depends on a countrys level o
14、f development income per capitaWe estimate the following growth regression model: g i?XiOPtelOPfinlnGDP?lnGDPi gi?XiOPtelOPfinlnGDPi?lnGDPi where i g is the average annual growth rate of per capita GDP for country i, Xi a vector of conditioning variables, OPtel and OPfin measure openness in telecomm
15、unication and financial services trade, respectively, lnGDPi is the log initial level of per capita income, and, ?11 ,?15?and ?21, ?25?are the threshold level and parameter vectors to be estimated, respectively. The focus of this study, the indices OPtel and OPfin, are from Mattoo, Rathindran and Su
16、bramanian 2001. They are constructed to reflect a policy-based measure of liberalization of a countrys services sectors. They assume values from 1 to 9 telecommunication services and 1 to 8 financial services with higher values indicating greater openness. The construction of the telecommunications
17、index was based on market structure, the degree of foreign ownership and the existence of independent regulators. The financial openness index was based on market structure, foreign equity, and Dailamis 2000 capital control index that takes into consideration the coding of rules, regulations and adm
18、inistrative procedures that can affect capital flows. One value of each index is available for each country referring to observations during the early 1990s. The dependent variable is the average annual growth rate between 1990 and 2000 to reflect the availability of the two indexes during the early
19、 1990s. The vector of conditioning variables, X i , includes: the government consumption/GDP ratio, lagged investment rate, inflation rate, mean years of total schooling in the population above 15 a measure of human capital and the ratio exports plus import/GDP. These variables were selected as repr
20、esentative of conditioning variables in the growth literature e.g. Levine et al., 2000. Data for all these variables are from the World Development Indicators, except for mean years of schooling from the Barro-Lee database. All conditioning variables are decadal averages.3. Empirical results Results
21、 are in Table 1The first four columns 1-4 show least squares estimates without a threshold. Column 1 is our basic growth specification. An issue of concern is that our two service trade indexes may be correlated with the general degree of openness and, thus, may proxy for the effects of overall trad
22、e openness. For this reason, the second specification column 2 includes an indicator of overall trade openness the ratio of imports plus exports to GDP. The third specification omits the investment rate to investigate the channels physical capital accumulation or productivity via which service openn
23、ess influences growth .The final specification column4 omits both the investment rate and overall trade opennessThe estimates show that greater openness in the telecommunication as well as financial services sector does not have a significant effect on growth.The conditioning variables have the expe
24、cted signs but, except for investment in physical capital, trade openness and government size in some specifications, are insignificant. Next, we employ the TR model. The bootstrapped Lagrange multiplier test statistics with 1000 replications are significant for all specifications p-values 0.011, 0.
25、009, 0.016 and 0.066. Therefore, income per capita splits the sample significantly. The least squares estimate of? is consistent across specifications:?$3310 for specifications 1-3 and$3150 for specification 4.1 For the first three specifications the threshold level divides the sample of 60 countrie
26、s into two subsamples 23 countries below and 37 above;2 for the final specification it divides the sample into a lower-income 21 and a higher-income group 39. Columns 1a-4a show parameter estimates for observations below the threshold and columns 1b-4b for observations above the threshold that corre
27、spond to the four specifications 1-4.The fit of the TR regressions as represented by R2 improves compared to least squares estimationFor countries below the threshold the relationship between openness in telecommunication services trade and growth is positive and highly significant; there is no evid
28、ence, however, of a significant relationship between openness in financial services and growth.For countries above the threshold the results are reversed: there is evidence of a positive and in two specifications significant relationship between openness in financial services trade and growth, but n
29、o such evidence for openness in telecommunication services trade.3 Overall trade openness has a positive and significant impact on growth for the high-income group and also a positive and marginally significant effect for the low-income groupWhen the investment ratio is excluded from the control var
30、iables in the final two specifications, the main result concerning the estimated coefficients for services openness for the two groups remains valid. We interpret this as evidence consistent with openness in services trade influencing the rate of economic growth both through physical capital accumul
31、ation and productivity growth. Finally, in contrast to the results for the complete sample, we note evidence favorable to conditional convergence for the two groups of countries separately, an observation consistent with the notion of convergence clubs. Because our results are based on the sample sp
32、lit via the TR model, we tested their robustness to alternative methods of sample splitting. One alternative is to introduce interaction terms between initial per capita income and the two indices of openness in services trade. The parameter estimates with t-statistics in parentheses for specificati
33、on 2 similar results hold true for the other specifications are:g0.029?0.023GovernmentSize?0.00094 Inflation+0.00107 Human Capital + 0.0456 Investment Rate + 0.00004 Trade Openness0.0045 Initial Income + 0.0174 OPTel ? 0.0147 OPFin0.0020 Initial Income OPTel These estimates confirm the results of th
34、e TR modelBoth interaction effects are highly significant, as are the estimated coefficients for both service openness indicators. Importantly, the coefficient on telecommunication services openness is positive and its interaction with initial income is negative, while the opposite is true for finan
35、cial services. This would indicate that telecommunication services openness has a positive impact on growth for low-income economies that diminishes as per capita income increases and vice versa for financial services, a conclusion that is in line with the TR model results. Based on these estimates,
36、 we can compute the level of income at which the impact of telecommunications openness on growth switches from positive to negative and vice versa for financial services openness: it is $6003 for telecommunication services openness and $2291 for financial services openness. Our findings on the growt
37、h effects of openness in telecommunication services trade for lower-income economies and financial services for higher-income economies brings into focus the issue of physical and human capital infrastructureWe conjecture that lower-income economies benefit from greater telecommunications openness b
38、ecause the degree to which investment in telecommunication services is successful in boosting productivity is linked to the availability of physical capital, a complementary factor. Efforts to liberalize telecommunication services trade by lower-income economies frequently elicit foreign direct inve
39、stment that contributes to raising productivity growth in the sector. For financial services, human capital a key complementary factor is in relative scarcity in lower-income economies and a factor that cannot be enticed from abroad, following liberalization.4Moreover, lower income countries do not
40、have the requisite institutional or regulatory structures to ensure the effective functioning of the financial sector as intermediary between savers and investors, something that greater openness of their financial sectors is unlikely to remedy. One concern with our results is the possible endogenei
41、ty of the two service openness indices. Several points are worth noting here. First, as indicated earlier, the service openness indices are based on observations largely from the first half of the 1990s while the growth rate is the average over the 1990-2000 decadeSecond, the correlation coefficient
42、 between the two service openness measures and the average growth rate is low: 0.11 for OPTel and 0.02 for OPFin both insignificant. Finally, we attempted to instrument for the service openness indices with the lagged growth rate but it turned out to be a poor instrument:the regression of each of th
43、e two service indicators on the lagged 1980-1990 growth rate yielded a positive but insignificant coefficient in both cases t-ratio 1.13 and 1.08 and the R2 was very low for both regressions approximately 0.02 in both cases. Given the way our service openness indices were measured, the low correlati
44、on coefficient between the two service openness indicators and growth, the persistence of growth rates across time a stylized fact and the low predictive power of the lagged growth rate for the service openness indicators, we do not expect endogeneity to be an important source of bias for our result
45、s.4. Conclusion This study aims to measure the effect of openness in telecommunication and financial services on economic growth for countries at different stages of economic development. We look at whether the impact of greater openness in these services on growth is different for low- and high-inc
46、ome countries. We find evidence of a positive and significant relationship between openness in telecommunication services and growth for countries with income per capita below an endogenously determined threshold level and no evidence of a significant relationship for countries above the same thresh
47、old. Concerning the financial services sector, the results are reversed: there is evidence of a positive and significant relationship between openness and growth for countries with income per capita above the threshold and no significant relationship for countries below the threshold. References1“Ca
48、nyon or Mirage, 2004, The Economist, January 22.2Dailami. M.2000, “Financial openness, democracy and redistributive policy, World Bank Institute.3Dollar. D., 1992, “Outward-oriented developing economies really do grow more rapidly: evidence from 95 LDCs 1976-1985, Economic Development and Cultural C
49、hange 40, 523-544.4Frankel. J, and D. Romer, 1999, “Does trade cause growth? American EconomicReview 89, 379-399.5Hansen, B.E., 2000, “Sample splitting and threshold estimation, Econometrica 68, 575-603.6Levine, R., N. Loayza and T. Beck, 2000, “Financial intermediation and growth: Causality and causes, J