統計コミュニティは統計不正にどう対応したか: 毎月勤労統計調査問題における政府・専門家・非専門家のはたらき

In Japan, we have recently witnessed several cases of long-overlooked misconduct in official statistics. The Japanese statistics community, which comprises people who are interested and skilled in statistics and capable of reading Japanese, does not appear to have made a remarkable contribution to detecting such misconduct. To discover the reason, this study focused on the incorrect method for estimating the number of workers introduced in 2018 to the Monthly Labour Survey by the Ministry of Health, Labour and Welfare (MHLW). On the basis of literature published since 2018, the author revealed that both inspecting committees in the government and external statistical experts overlooked indications of method alteration as well as the biases resulting from it, because they worked only on documents and explanations prepared by the MHLW. The finding suggests that, with the exception of a few non-expert citizens in the periphery, the Japanese statistics community lacks data-based criticism. At the core of the community, statistical experts uncritically and unjustifiably approve what government statisticians contend, without conducting original research on publicly available data.