arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1509.01482econ.GNq-fin.EC
收藏

衡量国家和产品的经济复杂性:应使用哪种指标?

Measuring economic complexity of countries and products: which metric to use?

University of Fribourg(弗里堡大学)

Manuel Sebastian Mariani, Alexandre Vidmer, Matus Medo, Yi-Cheng Zhang

AI总结:

本文比较了两种经济复杂性指标在国家和产品排名中的表现,发现Fitness-Complexity指标更优,并探讨了其通用化版本的改进效果。

摘要:

评估国家经济及其与全球市场产品的关系是经济学中的核心问题,对理论理解和实际应用如政策制定和金融投资规划都有深远影响。最近的经济复杂性方法旨在根据经验观察,即最具有竞争力的国家拥有多样化出口,而发展中国家只出口少量低质量产品——通常由其他国家出口。两种不同的指标,即Fitness-Complexity和反射方法,已被提出用于测量经济复杂性框架中的国家和产品得分。我们使用国际贸易数据和最近的排名评估措施,定量比较了这两种指标在根据网络重要性对国家和产品进行排名的能力。结果表明,Fitness-Complexity指标在产品和国家的排名中均优于反射方法。我们还研究了Fitness-Complexity指标的通用化版本,并显示只要输入数据可靠,它可以产生改进的排名。

英文摘要:

Evaluating the economies of countries and their relations with products in the global market is a central problem in economics, with far-reaching implications to our theoretical understanding of the international trade as well as to practical applications, such as policy making and financial investment planning. The recent Economic Complexity approach aims to quantify the competitiveness of countries and the quality of the exported products based on the empirical observation that the most competitive countries have diversified exports, whereas developing countries only export few low quality products -- typically those exported by many other countries. Two different metrics, Fitness-Complexity and the Method of Reflections, have been proposed to measure country and product score in the Economic Complexity framework. We use international trade data and a recent ranking evaluation measure to quantitatively compare the ability of the two metrics to rank countries and products according to their importance in the network. The results show that the Fitness-Complexity metric outperforms the Method of Reflections in both the ranking of products and the ranking of countries. We also investigate a Generalization of the Fitness-Complexity metric and show that it can produce improved rankings provided that the input data are reliable.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑