In 2009, the Institute of Public and Environmental Affairs (IPE) and the U.S. Natural Resources Defense Council (NRDC) jointly developed the Pollution Information Transparency Index (PITI) report, which is used to evaluate the EID situation of 113–120 local governments in China. It has experienced ten years of development up to now. Specifically, IPE and NRDC evaluate the situation of all the mediums set by local government to conduct the process of EID, that is, the process of obtaining the EI from polluting organizations, and then processing it through the platform and eventually disseminating to other stakeholders. Note that the cities or provinces included in the report were rated as national key environmentally friendly cities or regions because of good environmental performance, but some of them failed to perform as well as their EID showed, which have been detected by the IPE and NRDC. This phenomenon can be deemed as GLG caused by EID. Thereby, the cases from the PITI quite meet the demands of this research. In addition, a large number of scholars have carried out EID-related research by using the PITI report (e.g., [
23,
45,
46]), and the report has been widely recognized by governments, Non-Governmental Organizations (NGOs), consulting companies, and educational and research institutions (e.g., [
47]). The cases provided in this report are therefore authoritative and representative. Therefore, the cases in the PITI report become the research materials.
Although there are hundreds of related cases in the ten-year PITI reports, in order to improve the quality of the analysis and ensure the accuracy of the results, the paper selected the cases according to the following screening criteria: (1) They are derived from PITI reports after 2014 to ensure they are fresh; (2) The selected contents are credible and can be verified through various official mediums to ensure their authenticity; (3) They spread across different provinces and regions to ensure their universality; (4) They should also be in English which can benefit more audiences, thus the 2019 PITI report was excluded because of having no English version; (5) The cases are comprehensively described in PITI reports which can minimize the authors’ subjective bias. This work finally selected 20 satisfied cases from the reports.
The research team categorized the human causes from the 20 cases into each EID stage. In order to ensure the accuracy of classification, this article determined the following four steps to classify the human-caused risks in each stage. First of all, the confirmation of human-caused risks in each stage was based on factual evidence, and all information came from the traceable PITI report, which ensures the authenticity of the classification. Secondly, for the judgment of whether it is a human-caused risk, this research referred to the research experience of existing research to make the classification more convincing and scientific. The specific considerations were as follows. There are many methods for the classification of human-caused risks (e.g., [
48,
49,
50]). Among them, Reason [
50], the proponent of the Swiss cheese model, divided human-caused risks into four categories, namely “slips, lapses, mistakes, and violations.” Reason [
33] further pointed out that these human-caused risks stem from two aspects. One is due to abnormal psychological processes. The specific manifestations are: “forgetfulness, inattention, insufficient motivation, negligence and recklessness” [
38]. The second is due to errors in all aspects of a system that are operated by people (such as insufficient equipment reserves, poor design, poor supervision, manufacturing defects, shielding a shortcoming, insufficient training, poor automation, inappropriate procedures, or unclear definitions) [
38]. Scholars have used the above-mentioned classification of human-caused risks to carry out research in different fields (e.g., [
40]), which shows that these classifications have considerable feasibility and theoretical basis. Accordingly, this research regards the low data-collection rate, no timely EID, the problems of platform setting, and other issues exposed in the case as human-caused risks. Specifically, the low rate of data collection may be due to the above-mentioned insufficient motivation, poor supervision, and insufficient training. No timely EID is likely to be caused by a variety of human-caused risks such as negligence, poor automation, poor supervision, and inappropriate procedures. In addition, the problems of platform setting may also be caused by human factors such as poor design and poor automation. In view of the fact that the human-caused factors that lead to the GLG identified in this study involve more, the authors will not repeat the process of the analysis one by one here. Thirdly, after identifying human-caused risks based on the above criteria, this article needs to classify and summarize which stage of EID these man-made risks belong to. As explained in the section of methodology, EID is actually the whole process of collecting environmental information from polluting organizations, establishing environmental information dissemination media, and then disseminating it to other stakeholders by local governments. At different stages, there are obvious differences in the EID behavior of local governments. This article generally uses the keyword “collection” or synonymous as the standard when identifying the human-caused risks at the stage of collecting EI from pollution sources by local governments. For the EI media stage, it is concerned about the “platform” of information disclosure that is built by local governments. For the EI dissemination stage, it is concerned with the information disclosed by the government. Fourthly, in order to make the classification of this article more accurate, this study also adopted the expert discussion method. That is, with the help of well-trained assistants, professionals in this field independently match the review content one by one, then negotiate the differences until the opinions are fully agreed.