Waste decomposition analysis in Japanese manufacturing sectors for material flow cost accounting

Michiyuki Yagi, Katsuhiko Kokubu

Research output: Contribution to journalArticlepeer-review

11 Citations (Scopus)


From the perspective of material flow cost accounting (MFCA), which treats both material and financial flows within a company, this study proposes a corporate waste decomposition model to investigate the effects of material and financial factors on corporate waste generation. The proposed model decomposes waste into the material loss (waste ratio of raw materials [WRMat]), raw material-to-cost ratio (RtCR; material use efficiency), cost-to-sales ratio (or COGSR), total asset turnover ratio (TATR), leverage, and total equity. As an application, the waste decomposition analysis is performed using the log-mean Divisia index (LMDI) method, and 125 listed firms in 5 Japanese manufacturing sectors from 2010 to 2015 are analyzed. The LMDI results show that the RtCR, the most crucial term in MFCA, had the largest effect on increases in waste generation as of 2015; however, this effect is not so robust among sectors over the years, implying that MFCA is valid mainly for specific companies/sectors or years. Also, corporate environmental burdens (waste and carbon emission) are likely to be correlated negatively with leverage and positively with total equity in the models, implying that the financial and stock markets have an essential role in deciding corporate environmental burdens.

Original languageEnglish
Pages (from-to)823-837
Number of pages15
JournalJournal of Cleaner Production
Publication statusPublished - Jul 1 2019
Externally publishedYes

All Science Journal Classification (ASJC) codes

  • Renewable Energy, Sustainability and the Environment
  • Environmental Science(all)
  • Strategy and Management
  • Industrial and Manufacturing Engineering


Dive into the research topics of 'Waste decomposition analysis in Japanese manufacturing sectors for material flow cost accounting'. Together they form a unique fingerprint.

Cite this