Data for: Export sophistication and economic performance, new evidence using TiVA database

The data allow to investigate the relationship between export sophistication and economic performance for 64 countries over 2005-2015 period, based on Hausmann, Hwang and Rodrik (2007). PRODY and EXPY measures are computed using domestic value-added exports available from TiVA dataset instead of gross exports. TiVA dataset covers 35 sectors including 21 manufacturing and 14 services sectors, which allows to measure the impact of goods and services on income, alike. Other variables are gathered from different datasets. A dynamic panel GMM approach is followed. Income ratio defined as lnGDPpc/lnEXPY is employed as the dependent variable. Explaining variables include economic structure, technological content of exports, and TiVA new variables including backward and forward linkages variables. Strong evidence of the positive effect of manufacturing sector on countries’ economic performance is found. Weak evidence has been provided in favor of exports led growth hypothesis when taking conventional exports data into account, with the exception of high tech. and ICT exported goods, which have strong positive and significant effect on income. Relying on TiVA new indicators give new insights into countries GVCs participation gains. Thus, backward linkages seem to have an important role given their positive and significant effect on income, either sourced from commodities or services activities. Forward linkages seem to have mixed effects, depending on the end use of the exported domestic value-added, playing a prominent income role when domestic value-added is reimported, embodied in foreign final demand or when re-exporting intermediate imports as share of intermediate imports, suggesting that countries should not take GVCs’ benefits for granted.本数据集基于Hausmann、Hwang与Rodrik（2007）的研究框架，涵盖2005至2015年间64个国家的样本数据，可用于探究出口复杂度与经济绩效之间的关联。研究采用TiVA（Trade in Value Added，增加值贸易）数据集提供的国内增加值出口数据替代总出口数据，以此计算PRODY与EXPY指标。TiVA数据集覆盖35个行业门类，其中制造业21个、服务业14个，可同步测算商品与服务对国民收入的影响。其余变量来自多套公开数据集。本研究采用动态面板广义矩估计（GMM）方法进行实证分析。以lnGDPpc/lnEXPY定义的收入比值作为被解释变量，解释变量涵盖经济结构、出口技术含量，以及TiVA数据集新增的后向关联与前向关联等指标。实证结果表明，制造业对国家经济绩效存在显著正向推动作用。当采用传统出口数据进行检验时，除高技术产品与ICT（Information and Communication Technology）出口品外，出口导向型增长假说仅获得微弱的实证支持；而高技术与ICT出口品对国民收入具有显著且正向的影响。借助TiVA数据集的新型指标，可深入剖析各国参与全球价值链（GVCs，Global Value Chains）的收益。研究发现，后向关联指标对国民收入存在显著正向影响，无论其源于大宗商品还是服务活动，因此后向关联具有重要作用。前向关联的影响则呈现异质性，取决于出口国内增加值的最终用途：当国内增加值被重新进口并体现在外国最终需求中，或以中间进口再出口份额的形式体现时，前向关联对国民收入具有显著的促进作用。这表明各国不应将全球价值链的收益视为理所当然。