基于时间序列相似性的经济体出口发展模式性程度分析

doi:10.3969/j.issn.1004-9479.2025.11.20240291

摘要/Abstract

摘要：

基于覆盖世界217个经济体的96个产品类别的国际贸易数据，本文采用可比较性更优的相对性比较优势来描述各个经济体在各产品类别上的出口优势性，并用其构建各个经济体的出口发展路径，计算了各经济体之间在出口发展路径上的相似性程度。在此基础上，依次设定各经济体作为基准经济体，通过回归分析来挖掘其他各个经济体相对基准经济体的出口发展路径相似性程度与其自身经济增长速度的关系，并根据该回归分析所得的决定系数和回归系数来定义基准经济体的出口发展模式性程度指标。分析发现，以一部分经济体为基准时，通过出口发展路径相似性程度可以精准地描述各个经济体的经济增长速度；基于决定系数的出口发展模式性程度指标与经济体的人口规模呈现显著负相关，而基于回归系数一致性的出口发展模式性程度指标能够良好地解释各经济体在经济复杂性水平与经济适应度水平上的差异。这些发现表明可能存在一条对经济发展有着更高解释性的出口发展路径，这对挖掘各经济体的出口发展差异、理解国际贸易与经济发展之间的关系提供了新的认知角度。

关键词: 出口发展模式性程度, 出口发展路径, 相对性比较优势, 时间序列相似性分析, 相关性分析

Abstract:

This article focuses on the degree of an economy's export development pattern, which refers to the referentiality of an economy's export development pattern in promoting economic development in other economies. It proposes a set of metrics that leverage time-series similarity analysis to measure this degree from various perspectives and to explain economic development of economies. Using international trade data that covers 217 economies and 96 product categories, we employ relative comparative advantage as the measure of export advantage of an economy in each product category, owing to its better comparability. And then the export development paths of each economy are constructed based on the series of relative comparative advantage over the years. Subsequently, we calculate the similarity in export development paths between economies. By setting each economy as the benchmark economy in turn, regression analysis is used to explore the relationship between the similarity in export development paths relative to the benchmark economy and the economic growth rate of other economies. The degree of the benchmark economy's export development pattern is then measured based on coefficient of determination and regression coefficients derived from the regression analysis. For some benchmark economies, the economic growth rates of other economies can be accurately explained by the similarity in their export development paths. The metric measuring the degree of export development pattern, based on the coefficient of determination, reveals a significant negative correlation with the population size of economies. Furthermore, another type of metric for measuring the degree of export development pattern, assessed through the consistency of regression coefficients, can effectively explain the differences observed in the Economic Complexity Index and Economic Fitness among various economies. These findings imply the existence of a potential export development path that provides greater explanatory power for economic development, offering new cognitive perspectives for exploring the differences in export development among economies and understanding the relationship between international trade and economic development.

Key words: degree of export development pattern, export development paths, relative comparative advantage, similarity analysis of time series, correlation analysis

何靓, 韩筱璞. 基于时间序列相似性的经济体出口发展模式性程度分析[J]. 世界地理研究, 2025, 34(11): 59-72.

Liang HE, Xiaopu HAN. Analysis of economies' degree of export development pattern based on the similarity of time series[J]. World Regional Studies, 2025, 34(11): 59-72.

图/表 9

图1 各经济体的出口发展路径（以产品类型9为例）注：图a是在IR -g空间中各经济体的变化轨迹（滑动窗口宽度w = 0.3）；图b为各经济体的IR 随年份的变化。彩色曲线和灰色曲线分别对应各代表性经济体和其他国家。

Fig.1 Export development paths of various economies （using product type 9 as an example）

图2 不同基准经济体下每个经济体实际 δg 和预测值 E （ δg ）之间的关系注：各子图的基准经济体分别为中国（CHN）、美国（USA）、法国（FRA）、韩国（KOR）、俄罗斯（RUS）和巴西（BRA）。彩色数据点表示代表性经济体，其他数据点代表除代表性经济体以外的其他国家，虚线表示拟合直线；以下各图同。

Fig.2 The relationship between the actual δg and the predicted value E（δg） for each economy under different benchmark economies

图3 三类经济体在以帕劳（PLW）为基准经济体时所得的 δg 预测值和实际值的关系注：三个星号表示统计显著性P < 0.01。

Fig.3 Relationship between the predicted and actual values of δg obtained by three categories of economies with Palau as the benchmark economy

图4 各经济体的 R2 与人口规模〈log10p 〉之间的相关性

Fig.4 Correlation between R2 and population size 〈log10p〉 for each economy

表1 回归系数均值〈 q 〉的绝对值高于1.0×10-4 的产品类别

Tab.1 Product categories the absolute value of the mean regression coefficient 〈q〉 is higher than 1.0×10-4

编码	产品类别	〈q〉(×10^-4)	编码	产品类别	〈q〉(×10^-4)
1	活动物	1.15	43	毛皮、人造毛皮及其制品	-1.17
10	谷物	-1.63	49	书籍、报纸、印刷图画及其他印刷品，手稿、打字稿及设计图纸	-2.26
14	编结用植物材料、其他植物产品	-1.03	51	羊毛、动物细毛或粗毛，马毛纱线及其机织物	-1.06
20	蔬菜、水果、坚果或其他植物制品	-2.43	61	针织或钩编的服装及衣着附件	-2.53
21	杂项食品	-1.08	64	鞋靴、护腿和类似品及其零件	-1.24
25	盐，硫磺，泥土及石料，石膏料、石灰及水泥	1.07	74	铜及其制品	1.40
30	药品	-2.63	75	镍及其制品	1.07
31	肥料	-1.02	76	铝及其制品	1.82
35	蛋白类物质、改性淀粉、胶、酶	1.22	83	贱金属杂项制品	-1.99
38	杂项化学产品	1.21	84	核反应堆、锅炉、机器、机械器具及其零件	-2.09
40	橡胶及其制品	-1.95	87	车辆及其零件、附件、铁道及电车道车辆除外	-1.28
42	皮革制品，鞍具及挽具，旅行用品、手提包及类似容器，动物肠线（蚕胶丝除外）制品	-1.33	92	乐器及其零件、附件	-1.11

图5 回归系数一致性 ρE 与 δg 、〈 KE 〉、〈log10F 〉之间有着最强关系的实例注：面板a）、b）、c）的基准经济体分别为玻利维亚（BOL）、美国（USA）和丹麦（DNK）。

Fig.5 An example of the strongest relationship between the consistency of regression coefficients ρE and δg，〈KE 〉，〈log10F〉

图6 ρpδg、ρpK和ρpF与人口规模〈log10p 〉的关系

Fig.6 The relationship between ρpδg、ρpK andρpF and population size 〈log10p〉

图7 回归系数一致性ρpδg对 δg 的解释性与 δg 的相关性

Fig.7 Explanatory effect of consistency of regression coefficients ρpδg?on δg and correlation with δg

图8 ρpK和ρpF与ECI水平〈 KE 〉和适应度水平〈log10F 〉的相关性

Fig.8 Correlation between ρpK and ρpF with ECI level 〈KE 〉 and fitness level 〈log10F〉

参考文献 44

[1]	HAUSMANN R, HIDALGO C. The network structure of economic output. Journal of Economic Growth, 2011, 16(4): 309-342.
[2]	LIN J. New structural economics: A framework for rethinking development. The World Bank Research Observer, 2011, 26(2): 193-221.
[3]	SUTTON J, TREFLER D. Capabilities, wealth, and trade. Journal of Political Economy, 2016, 124(3): 826-878.
[4]	EATON J, KORTUM S. Technology, geography, and trade. Econometrica, 2002, 70(5): 1741-1779.
[5]	CALIENDO L, PARRO F. Estimates of the trade and welfare effects of NAFTA. The Review of Economic Studies, 2015, 82(1): 1-44.
[6]	DIXIT A, NORMAN V. Theory of International Trade: A Dual, General Equilibrium Approach. Cambridge University Press, 1980.
[7]	HU Y, FAN Z, LIN J, et al. A reasoning of economic complexity based on simulated general equilibrium international trade model. Journal of Physics: Complexity, 2023, 4(3): 035003.
[8]	BALASSA B, NOLAND M. "Revealed" comparative advantage in Japan and the United States. Journal of International Economic Integration, 1989, 4(2): 8-22.
[9]	HAUSMANN R, KLINGER B. The structure of the product space and the evolution of comparative advantage. CID Working Paper No.146, Cambridge, MA: Harvard Kennedy School, Harvard University, 2007.
[10]	LEAMER E E. Sources of Comparative Advantage: Theory and Evidence. Cambridge:MIT Press, 1984.
[11]	LALL S. The Technological structure and performance of developing country manufactured exports, 1985-98. Oxford Development Studies, 2000, 28(3): 337-369.
[12]	HIDALGO C A, KLINGER B, BARABÁSI A L, HAUSMANN R. The product space conditions the development of nations. Science, 2007, 317(5837): 482-487.
[13]	HAUSMANN R, HIDALGO C. The building blocks of economic complexity. Proceedings of the National Academy of Science of the United States of America, 2009, 106(26): 10570-10575.
[14]	HAUSMANN R, HIDALGO C. The network structure of economic output. Journal of Economic Growth, 2011, 16: 309-342.
[15]	TACCHELLA A, CRISTELLI M, CALDARELLI G, et al. A new metrics for countries' fitness and products' complexity. Scientific Reports, 2012, 2(1): 723.
[16]	HAUSMANN R, HIDALGO C, BUSTOS S, et al. The Atlas of Economic Complexity: Mapping Paths to Prosperity. Cambridge: MIT Press, 2014.
[17]	STOJKOSKI V, UTKOVSKI Z, KOCAREV L. The impact of services on economic complexity: Service sophistication as route for economic growth. PloS One, 2016, 11(8): e0161633.
[18]	PONCET S, WALDEMAR F. Economic complexity and growth. Revue Economique, 2013, 64(3): 495-503.
[19]	DOMINI G. Patterns of specialisation and economic complexity through the lens of universal exhibitions. Laboratory of Economics and Management (LEM), 2019, 20: 1855–1900.
[20]	OURENS G. Can the Method of Reflections help predict future growth? FCS-DE, 2012-17.
[21]	CHÁVEZ J, MOSQUEDA, GÓMEZ-ZALDÍVAR M. Economic complexity and regional growth performance: Evidence from the mexican economy. Review of Regional Studies, 2017, 47(2): 201-219.
[22]	MEALY P, FARMER J, TEYTELBOYM A. Interpreting economic complexity. Science Advances, 2019, 5(1): eaau1705.
[23]	SBARDELLA A, PUGLIESE E, PIETRONERO L. Economic development and wage inequality: A complex system analysis. PloS One, 2017, 12(9): e0182774.
[24]	FAWAZ F, RAHNAMA-MOGHADAMM M. Spatial dependence of global income inequality: The role of economic complexity. The International Trade Journal, 2019, 33(6): 542-554.
[25]	CHU L, HOANG D. How does economic complexity influence income inequality? New evidence from international data. Economic Analysis and Policy, 2020, 68: 44-57.
[26]	ZHU S, YU C, HE C. Export structures, income inequality and urban-rural divide in China. Applied Geography, 2020, 115: 102150.
[27]	LEE C, WANG E. Economic complexity and income inequality: Does country risk matter? Social Indicators Research, 2021, 154(1): 35-60.
[28]	BREITENBACH M C, CHISADZA C, CLANCE M. The Economic Complexity Index (ECI) and output volatility: High vs. low income countries. The Journal of International Trade & Economic Development, 2022, 31(4): 566-580.
[29]	SARACCO F, DI CLEMENTE R, GABRIELLI A, et al. Detecting early signs of the 2007–2008 crisis in the world trade. Scientific Reports, 2016, 6(1): 30286.
[30]	CRISTELLI M, TACCHELLA A, CADER M, et al. On the predictability of growth. World Bank Policy Research Working Paper, 2017: 8117.
[31]	TACCHELLA A, MAZZILLI D, PIETRONERO L. A dynamical systems approach to gross domestic product forecasting. Nature Physics, 2018, 14(8): 861-865.
[32]	CRISTELLI M, TACCHELLA A, PIETRONERO L. The heterogeneous dynamics of economic complexity. PloS One, 2015, 10(2): e0117174
[33]	MARIANI M, REN Z, BASCOMPTE J, et al. Nestedness in complex networks: observation, emergence, and implications. Physics Reports, 2019, 813: 1-90.
[34]	HOERL A, KENNARD R. Ridge regression: applications to nonorthogonal problems. Technometrics, 1970, 12(1): 69-82.
[35]	MARQUARDT D, SNEE R. Ridge regression in practice. The American Statistician, 1975, 29(1): 3-20.
[36]	KIBRIA B M G. Performance of some new ridge regression estimators. Communications in Statistics-Simulation and Computation, 2003, 32(2): 419-435.
[37]	MCDONALD G. Ridge regression. Wiley Interdisciplinary Reviews: Computational Statistics, 2009, 1(1): 93-100.
[38]	TSIGLER A, BARTLETT P. Benign overfitting in ridge regression. Journal of Machine Learning Research, 2023, 24(123): 1-76.
[39]	LI R, DONG L, ZHANG J, et al. Simple spatial scaling rules behind complex cities. Nature Communications, 2017, 8: 1841.
[40]	WANG C, WANG L, XUE Y, LI R. Revealing spatial spillover effect in high-tech industry agglomeration from a high-skilled labor flow network perspective. Journal of Systems Science & Complexity, 2022, 35: 839-859.
[41]	易先忠, 欧阳峣. 大国如何出口:国际经验与中国贸易模式回归. 财贸经济, 2018, 39(3): 79-94.
	YI X, OUYANG Y. How should large economies export: international experience and rebuilding of China's export model. Finance & Trade Economics, 2018, 39(3): 79-94.
[42]	UTTERBACK J, SUÁREZ F. Innovation, competition, and industry structure. Research Policy, 1993, 22(1): 1-21.
[43]	BOYKO C, COOPER R. Density and decision-making: Findings from an online survey. Sustainability, 2013, 5(10): 4502-4522.
[44]	SPENGLER J. Population and Economic Growth.//Population Growth. London: Routledge, 2017: 59-69.