Wataru Kobayashi, Director of Chiba University of Commerce Research Center for Economics

Aim of the special feature in "CUC View & Vision No. 53"

In Issue 51 of this magazine (published on March 31, 2022), we introduced how model analysis, which is considered one of the standard analytical methods in economics, is handled in other fields of social science. In this special feature, as a sequel to that issue, researchers in each field will explain how data analysis methods are being deployed in various fields of social science.

The first article introduces examples of data analysis in economics. Economics studies how people make decisions and how resources are allocated as a result. However, it is necessary to use data to verify whether theoretical conclusions derived under certain assumptions apply to actual economic phenomena. For example, economics textbooks say that minimum wage hikes lead to an increase in unemployment. This is because it is believed that minimum wage hikes lead to an increase in labor costs for companies, which in turn suppresses employment. One way to verify this hypothesis is to obtain data on employment and the number of unemployed in countries and regions where the minimum wage has actually been raised, and to confirm how employment and unemployment changed before and after the hike. However, even if the amount of employment after the minimum wage hike is higher than before, it is possible that factors other than the minimum wage are influencing the situation, such as the economy simply improving at that time. Taking such things into consideration, various efforts have been made to extract causal relationships, and several research examples are introduced here, mainly from research that has won the Nobel Prize in Economics.

The second article provides an overview of processed statistics in macroeconomic analysis, focusing on the System of National Accounts (SNA) and Gross Domestic Product (GDP). There are various indicators referred to in macroeconomics, but GDP is characterized by being a processed statistic (secondary statistics) created through processing of basic statistics (primary statistics). In other words, new data is created by processing observed data. In addition to real transactions, the value of GDP also includes activities such as imputed rent services, which do not actually involve monetary transactions. On the other hand, GDP has various problems, and alternative indicators have been devised, but GDP is still considered one of the most important indicators in macroeconomic analysis. Here, we consider the background to the creation of such statistics and the reasons why their position is still unshakable.

The third paper examines the methods and possibilities of data analysis in management accounting research. Management accounting is a field of accounting that deals with internal management of organizations, and is closely related to economics and business administration in the sense that it studies corporate decision-making. In this paper, we outline three conventional research methods: archival, experimental, and survey, and then consider the possibility of applying neuroscience knowledge to data analysis as a new research method. Archival is a method of collecting and analyzing information disclosed by organizations to third parties, and focuses on the analysis of published financial data. Experiments are a method of giving subjects specific tasks and information, and collecting and analyzing data from their decisions and actions. And surveys are a method of conducting questionnaire surveys of the target population, collecting and analyzing information. In contrast, neuroscience aims to measure the activity of the human brain, and it is said that it has the potential to elucidate areas that have been a black box in previous research.

The fourth article outlines research examples on the occurrence of human errors in production management. Manufacturing companies need a BOM (Bill of Materials) that hierarchically shows the components that make up a product. However, when new products are introduced and design changes occur frequently, it becomes difficult to spend enough time creating the BOM. In addition, in recent years, product specifications have become more complex, and the contents of the BOM have become increasingly complex as well. As a result, errors in creating BOMs at manufacturing sites are occurring frequently, and research is being conducted to identify factors that affect worker behavior in order to devise effective measures to reduce human errors. Here, we introduce two such studies that focus on the review work of the created configuration master (part of the BOM) and the work of creating the configuration master itself. In both cases, data is obtained through experiments and hypotheses are verified using a statistical method called analysis of variance.

The fifth article focuses on the organizational management of neighborhood associations and town associations from the perspective of social capital theory, and examines the sustainable management of local communities required in the post-COVID era. In this regard, the use of mixed research methods, which combine quantitative and qualitative research, is emphasized. Typical combinations of these methods are to conduct quantitative research such as questionnaire surveys first, and then conduct qualitative research such as interviews to deepen understanding of the results, and to conduct qualitative research first and then conduct quantitative research to verify hypotheses set based on the results. As an example of the former, the article outlines the results of a questionnaire survey conducted on citizen groups in Kamagaya City, Chiba Prefecture, and interviews conducted on some of these groups, and as an example of the latter, the article outlines the results of an interview survey conducted on neighborhood associations in Katsushika Ward, Tokyo, and a questionnaire survey conducted on residents of Tokyo's 23 wards to verify the hypotheses derived from the survey.