期刊文献+

大数据如何改变经济学研究范式? 被引量:71

How Is Big Data Changing Economic Research Paradigms?
原文传递
导出
摘要 本文首先从经济学视角探讨大数据给经济学实证研究所带来的范式变革,包括从理性经济人到非完全理性经济人,从孤立的经济人到互相关联的社会经济人,从代表性经济人到异质性经济主体,以及从经济分析到经济社会活动的系统分析。然后,从方法论视角讨论大数据给经济学实证研究方法所带来的变革,包括从模型驱动到数据驱动,从参数不确定性到模型不确定性,从无偏估计到有偏估计,从低维建模到高维建模,从低频数据到高频甚至实时数据,从结构化数据到非结构化数据,从传统结构化数据到新型结构化数据,以及从人工分析到智能分析等。大数据引起的经济学研究范式与研究方法变革,正在深刻重塑经济学发展方向,不但加强了经济学实证研究范式的趋势,而且还进一步突破了现代西方经济学的一些基本假设的局限性,使经济学研究日益呈现出科学化、严谨化、精细化、多元化(跨学科)与系统化的趋势,并且与社会科学其他领域在方法论上日益趋同。中国大数据资源,为从中国经济实践中总结经济发展规律,从中国特殊性中凝练可复制的经济发展模式,从而构建具有深厚学理基础的原创性中国经济理论体系,提供了一个得天独厚的"富矿"。 With the emerging new generation of information technology revolution and the Fourth Industrial Revolution,massive interrelated and high-frequency Big data of microeconomic behaviors and activities are being generated and recorded,examples being data from mobile smartphones,satellites and sensors,scanning machines,digital business platforms,digital social medias,government and private institution websites,and digital libraries.The Big data revolution is profoundly changing the ways of human production and life as well as the research paradigms and methodologies in economics.In this paper we first investigate the paradigm shifts brought by Big data to the mainstream economic empirical research,particularly the relaxation of some fundamental assumptions,including from a rational economic agent to a non-completely rational economic agent,from an isolated economic agent to socially connected economic agents,from a representative economic agent to heterogeneous economic agents,and from economic analysis to the systematic analysis of econ-social behaviors and activities.These changes and shifts make economic modeling and analysis much closer to economic reality and are expected to offer new knowledge discoveries about and insights into the economy.Next,we discuss the methodological changes brought by Big data to economic empirical research,including from a model-driven approach to a data-driven approach,from focus on the impact of parameter estimation uncertainty to focus on the impact of model uncertainty,from the use of unbiased estimators to the use of regularized(biased)estimators,from a low-dimensional modeling strategy to a high-dimensional modeling strategy,from the use of low-frequency data to the use of high-frequency and even real-time data,from the use of structured data to the use of unstructured data(such as texts,graphs,photos,audio and video),from the use of traditional structured data to the use of new kinds of structured data(such as matrix data,functional data,interval-valued data and symbolic data),and from human data analysis to intelligence data analysis.Among many other things,Big data offers new information not available in traditional data,making it possible to conduct cutting-edge interdisciplinary research and to transform qualitative analysis to quantitative analysis.Leading examples include the textual regression analysis on the impact of investor sentiments on asset pricing due to the availability of text data,and analysis of heterogeneous distributional effects of macroeconomic policies and external shocks on different sectors or groups of economic agents due to availability of microeconomic high-frequency time series data.The revolution in economic research paradigms and methodologies,brought by Big data,is reshaping the directions of modern economics,especially strengthening the trend of economic empirical research and breaking the restrictions of some basic assumptions of modern Western economics to various extents.As a result,economic research becomes increasingly scientific,rigorous,elaborate,interdisciplinarily diverse,and systematically integrated,and is increasingly similar to other fields of social science in terms of the use of quantitative methodologies.Given the scale advantages of mega economy,the rapid development of digital economy,and the rich experience of government reforms and policy-makings,China’s Big data resources provide a unique"rich ore"for Chinese economists to build an innovative theoretical system of Chinese economics by summarizing China’s economic development rules and developing a replicable China’s economic development model from the past 70 years of Chinese economic practice,particularly from the past 40 years of economic reforms and open-door policies to the outside world.
作者 洪永淼 汪寿阳 Hong YongMiao;Wang Shouyang
出处 《管理世界》 CSSCI 北大核心 2021年第10期40-55,72,M0004,243,共19页 Journal of Management World
基金 国家自然科学基金委员会基础科学中心项目“计量建模与经济政策研究”(项目号:71988101) 国家自然科学基金专项项目“经济科学发展战略研究”(项目号:71940004)的资助。
关键词 大数据 文本分析 机器学习 研究范式 研究方法 反身性 Big data textual analysis machine learning research paradigms methodologies reflexivity
  • 相关文献

参考文献18

二级参考文献194

共引文献863

同被引文献1454

引证文献71

二级引证文献515

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部