Theoretical and empirical research for making and using Economic Statistics

经济统计数据对于正确分析真实世界经济非常重要。但在我们看来，似乎有很多实证研究机械地应用了几种统计方法，对统计数据存在的内在问题缺乏深入的了解。日本社会发生了翻天覆地的变化，这主要是由于快速的信息化和国际化。这给经济统计带来了严重的问题，可以认为它不能描述整个社会的情况。因此，我们需要调查和审视这些问题，并寻求解决这些问题的方法和经济统计的未来形象。造成这些问题的主要原因是经济统计的制定者和使用者之间没有密切的沟通。在认识到这是经济学和统计学领域的关键问题之一的基础上，我们的项目主要关注以下三点：(1)澄清目前统计系统…存在的问题更多的是在欧洲和亚洲国家，(2)从统计调查理论的角度重新审视全国人口普查和民意研究的具体问题，(3)澄清私营机构各种调查统计的现状。我们目前研究的主要观点可以概括为：(A)现有的经济统计存在许多问题，这些问题来自于传统的、比较僵化的分类体系(工业、职业和商品)。因此，我们需要一个反映社会“信息化”和“国际化”的新分类系统。(B)由于人们的隐私意识和不断变化的社会环境，给我们提供第一批原始数据的人口普查和抽样调查现在变得越来越困难。我们需要重新考虑整个统计系统，以适应这种环境。(C)从用户的角度来看，目前报告的统计数据太多，报告系统过于复杂，无法了解每项数据的特点和范围以及它们之间的相互关系。我们需要一些好的、通俗易懂的《经济统计指南》。(D)私人机构进行的抽样调查有许多突出的特点，覆盖范围广，官方统计数据有些不能涵盖和掌握实际情况，改进和发展适合分析瞬息万变的经济统计数据的统计技术是非常迫切的。较少

Economic statistics are very important in analysising the real world economy properly. But it seems to us that there are many empirical researches which applied several statistical methods mechanically without deep understanding about there inherent problems of statistical data. The Japanese society has been drastically changing mainly due to rapid informatization and internationalization. This causes serious problems in economic statistics, which can be thought as away of describing the situation of society as a whole. We need, therefore, to investigate and scrutinize such problems and to seek for means of their solutions and future image of economic statistics.The main reason for these problems is no intimate communication exits between makers and users of economic statistics. Based on the recognition that this is one of the key issues in the field of economics and statistics, our project mainly focuses on the next three points: (1) clarify the current problems of statistical system … More in European and Asian countries, (2) reinvestigate the specific problem of national census surveys and public opinion research from the point of view of statistical survey theory, and (3) clarify the current status of all kinds of the survey statistics of private organization.The main points which we have gotten from our research so far can be summarized in the followings:(a) The existing economic statistics are now suffering many problems, which came from the traditional and rather rigid system of classification (on industry, occupation and commodity). Therefore, we need a new classification system reflecting "informatization" and "internationalization" of the society.(b) The censuses and sample surveys, which give us the first original data, are now getting more difficult because of people's consciousness of privacy and changing social environment. We need to reconsider the statistical system as a whole so as to fit this environment.(c) From users' viewpoint, these days, too many statistics are now being reported and their reporting system is too complicated to understand the features and scope of each data and the interrelationship among them. We need some good "Guidebook of Economic Statistics" which is easily understandable.(d) The sample survey conducted by private organization has many prominent features; covering wide area, some of which official statistics fail to cover and grasping the actual conditions.It is very urgent for us to improve and develop statistical techniques suitable for the analysis on economic statistics which are drastically changing. Less