摘要
中国工业企业数据库是近年来国内外经济学界常用的企业级微观数据库,但从统计抽样调查的角度来看,该数据库存在着各种问题。为更好地了解该数据库可能出现的系统性误差与随机误差,本文使用1996-2013年数据对真实性、系统性误差及基层统计部门反映的各种数据问题,进行定量评估。结果发现:样本范围及统计口径的变动,缺失值较多、"回避规模以上"、"化整为零"等方面的数据问题,均会对数据产生一定的影响,并据此在行业代码整理、真实性检验、统计口径变更处理、统计偏误克服、通货膨胀平减等方面提出了相关建议。
Abstract:China's industrial enterprise database is one of the most commonly used microdatabase at enterprise level both for domestic and foreign economic scholars in recent years.From the perspective of statistical sampling survey,however,it's more or less problematic.In order to better understand possible systematic errors and casual errors of this database,this paper uses the data from 1996 to 2013 to make quantitative assessments about its authenticity,systematic errors and various data problems reflected by the grassroots statistics department.The results show that the fluctuations of sample range and statistical caliber,a large number of missing values,and the problems related to enterprise statistics personnel'filling and reporting such as "the avoidance of above designated size"and "breaking up the whole into parts"will produce objective impacts on the data.Accordingly,this paper proposes corresponding solutions and suggestions.
作者
陈林
Chen Lin(Business School,Beijing Normal University)
出处
《经济评论》
CSSCI
北大核心
2018年第6期140-153,共14页
Economic Review
基金
国家自然科学基金面上项目"竞争政策与准入规制的协调机制研究"(项目号:71773039)的资助
关键词
中国工业企业数据库
规模以上工业企业
工业统计报表
企业级微观数据:工业增加值
Chinese Industrial Enterprise Database
Industrial Enterprises above Designated Size
Industrial Statistical Reporting System
Micro Enterprise Data
Industrial Added Value
