5.1. What Do the Results Mean for Japan?
Application of Equation (2) yields an EEC of 14% and 6.1% for low-income and high-income households, respectively. Comparing energy costs from 2000 to 2014, the total economic burden placed upon low-income households is more severe than the burden on high-income households. Low-income households suffered an increase EEC from 9.4% (2000) to 14% (2014), while high-income housing only saw a relative increase of 5.3% (2000) to 6.3% (2014) [13
]. Greater expendable income means individuals/families in the high-income bracket could cut back on excessive use, but still have sufficient energy to maintain their lifestyle. Low-income families do not have the same luxuries (i.e., any cutback in energy consumption would result in a significant degradation to quality of life). Thus, low-income families are required to pay the, currently excessive, energy costs to sustain a manageable standard of living. These costs do not include the full cost of nuclear insurance.
Another analysis was compiled from data collected by the Japan Statistics Bureau, Ministry of Internal Affairs and Communications (MIAC). Similar to Ikarii’s study, this analysis compares similar metrics from 2003 to 2015, while household income is divided into quintile groups. The approximate segregation in the year income quintiles are as follows: quintile I: maximum 2,680,000 yen (~$26,000), quintile II: 2,680,000 to 3,960,000 yen (~$26,000–39,000), quintile III: 3,960,000 to 5,400,000 yen (~$39,000–53,000), quintile IV: 5,400,000 to 7,780,000 yen (~$53,000–76,000), quintile V: minimum 7,780,000 yen (~$76,000) [23
]. Figure 1
describes the trend of increased energy cost for low-income (quintile I) and high-income (quintile V) families for comparison to Ikarii’s study.
displays the entire deviation for all noted quintiles from 2003 to 2015. Thus, it is again apparent that the increase in energy cost is a further burden upon the low-income tiered households.
The values described earlier evaluate the relative cost implications as a result of a significant decrease in the application of nuclear power plants for nuclear generation after the Great East Japan Earthquake. However, the figures do not yet account for the possible insurance cost(s) borne by Japanese energy providers and subsequently the consumers. By merging the calculated yearly electrical consumption costs per household (Table 8
) with Fukushima (Daiichi) insurance costs, a new valid EEC value can be determined. Relative to low-income and high-income households, the redistributed EEC values are larger, in all circumstances, compared to the 2010 pre-disaster metrics. Figure 2
compares the traditional model for EEC in 2010 vs. the EEC calculated with insurance costs per news source.
Determined EEC metrics suggest that a significant, if not illogical, amount of consumer funds would have to be utilized to properly support the required insurance premiums of nuclear power plants. As demonstrated by Figure 2
, for most low-income households, more than 100% of their expenditure would be needed solely to pay for electrical energy. In extreme cases, the true cost of the nuclear power plant generated electrical energy exceeds 250% of a household’s current total consumption expenditure. Further utilizing 2010 METI reported survey results, the total consumption expenditure for quintiles I, II, III, IV, and V was ~85%, 78%, 65%, 55%, and 48%, respectively, of their yearly income [12
]. Thus, significant increases will result in a significant portion, if not all, of a household’s income being dedicated to paying the electricity bill. The extent of this cost is such that all quintiles will be affected to the fullest extent (i.e., it would impact their quality of life with regard to monetary freedom).
5.2. The Price–Anderson Act (PAA) and U.S. Government Indirect Nuclear Subsidy in Perspective
The academic literature currently provides limited thorough economic analyses of liability limits under the Price-–Anderson Act (PAA). To date, the most complete and comprehensive analysis was performed by Dublin and Rothwell in 1990 [23
]. The study [24
] aimed to compare the insurance premiums nuclear power plant operators currently pay under the PAA to the premiums weathered if facilities were responsible for full liability coverage in the absence of PAA limits. The work presented by Dublin and Rothwell is conservative (e.g., no health effects are included in the analysis). Thus, the estimated insurance subsidy per reactor was $60 million in 1982 [24
]. The 1988 amendments to the PAA increased the total provided coverage to $7 billion, which consequently reduces the subsidy per reactor to $22 million. In total, by 1988 the total subsidy available to the nuclear industry was $111 billion. In 1998, Heyes and Liston-Heyes published an article that modified the methodology applied by Dublin and Rothwell with regard to the insurance arrangements. The article suggests that Dublin and Rothwell’s method misinterpreted the terms on which the insurance arrangements are applied [25
]. The corrections presented reduce the reported subsidy per reactor by a factor of four and 10. Specifically, the Price–Anderson liability limit was $13.32 million before the 1988 amendments and $2.32 million after the amendments [25
]. However, the correction presented implies that the current risk assumptions are correct (i.e., the provided nuclear subsidy is adequate for the risk environment of a nuclear power plant). In a 2003 publication, Heyes acknowledges the fact that he is merely an economist with limited understanding of the working principles of a nuclear facility. As such, economic models by Dublin and Rothwell, later re-imagined by Heyes-Liston Heyes, may not accurately present the true consequence-probability model [26
]. While updating the 1988 estimate of $22 million per reactor to $33 million (2001 dollars) per reactor, Heyes suggests that the actual subsidy could be even higher. Heyes further describes an ideal situation in which a cost assessment, based on private industry nuclear knowledge, is readily available, however, such knowledge is not accessible. As a result of the lack of quantitative information, further economic analysis of nuclear power, as stated by Dublin, “depend[s] crucially on assumptions” [26
Thus, Heyes’s 2003 estimate of $33 million (2001 dollars) per reactor can be compared to the values calculated for the Fukushima (Daiichi), a six reactor facility here. In an extremely conservative comparison, utilizing The Guardian’s relatively low estimate of $20 billion in insurance costs, distributed over Japan’s 43 operational reactors yields ~$456 million per reactor in the total required insurance pool. Thus, even using the lowest estimate, which excludes the vast majority of the liability, of Fukushima (Daiichi)’s real costs the equivalent U.S. subsidy amounts to over ~$192.3 million per reactor. Alternatively, utilizing the highest and most comprehensive estimates (e.g., The Economist), the total U.S. subsidy amounts climbs to ~$4.4 billion per reactor. The field of nuclear insurance subsidy remains considerably under-investigated, and although this study represents the most comprehensive to date because it uses post-disaster data (with the limitations of the data discussed above) it is instructive to compare it to the other studies conducted on the topic, which have focused on the USA. In all sources utilized, including Forbes, The Guardian, The Economist, The Wall Street Journal, Reuters, and Bloomberg, the reported insurance costs greatly exceed the current PAA insurance pool of $121 million per reactor.
To further clarify the magnitude of the evaluated costs, consider if the Fukushima (Daiichi) disaster costs were determined as a baseline analysis to be compared to the current U.S. government PAA subsidy. A six-reactor facility similar to Fukushima (Daiichi) would have been provided in total $726 million for cleanup compensation beyond the first-tier $375 million. Although this is a large sum, it is apparent that the subsidy fails to adequately provide a sufficient backing to a nuclear disaster by ~(-)180% by comparing insurance backing vs. actual. In the lowest estimates, The Guardian
reports $20 billion in damages, which is ~$19 billion greater than the available PAA compensation that would be provided via the PAA insurance pool. In fact, the lowest reported cost of a nuclear facility with six reactors (Fukushima) is greater than the total insurance pool for the entire U.S. nuclear power plant fleet at $12.6 billion (utilizing 104 reactors). It has been claimed before that the building of a nuclear power plant is not possible without government involvement to cover accident liability [23
]. The economic analysis, with regard to the value of the government subsidy, has not accounted for the excessive costs associated with a nuclear disaster. Although rare in occurrence, nuclear accidents’ costs and subsequent damages are not fully accounted for, as is evident by the deficit of PAA support vs. the cost of a six-reactor catastrophe.
For nuclear energy to be a prominent energy source, the economic, safety, engineering, and climatology impacts must be fully understood prior to becoming a sustainable energy source [27
]. The aforementioned cost implications are not fully understood and as a result nuclear energy generation continues to be viewed by the public as an economically viable source. Nuclear insecurity must be reduced to prevent accidents like Chernobyl [28
] and prevent new national sacrifice zones [30
]. This means there is a need to reduce nuclear power risk factors, which include potential nuclear disasters due to accidents, negligence, poor design, natural disasters, and terrorism [32
] (as a prime target, but also for theft and the creation of ‘dirty bombs’). The public is aware of the risks associated with nuclear power plants, as evidenced by, for example, the willingness of Chinese consumers to pay more for electricity to stop their communities from housing a nuclear power plant [33
]. It is clear from this study, however, that, as seen in the past [34
], the real economic costs of nuclear power far exceed the competing sources of electrical generation. It is also clear that when the full costs of insurance are included with current nuclear power systems, they are not economical. Furthermore, a government-provided insurance guarantee removes the economic incentive for the nuclear industry to seek private insurers. As a result, taxpayers are at risk of being subjected to continuous payments during operation and potential nuclear disaster of any nuclear power plant. Worse than this, the low liability cap placed upon the nuclear industry may not incentivize the nuclear power plant facilities to invest in proper safety measures [35
]. Similarly, a liability cap of $75 million for offshore drilling could have been a contributing factor in the British Petroleum (BP) Gulf of Mexico oil disaster. BP will utilize $20 billion to correct this environmental disaster, which calls into question the economics of liability caps, including clean-up costs and safety incentives [36
5.3. Availability and Accountability of Post-Disaster Economic Reporting
The results shown above underscore the need for high-quality journalism on issues related to public hazards, economics, and energy policy. The significant variability, ±$194 billion, apparent in the sourced Fukushima (Daiichi) cost accounting is troubling. Understandably, each news agency utilizes its own sources in reporting. For example, Forbes [17
] used a secondary story about TEPCO [37
] and data from Japan’s reconstruction agency [38
] in the story. Other news providers like The Economist
and CNN did not provide direct links to sources, but, for example, quoted former employees of Japan’s Atomic Energy Commission or TEPCO employees. In addition, some news agencies did not specifically cite a source (or a method) for their estimate. Thus, within the given dataset the validity of the reported cost data must be questioned. Specifically, the standard deviation is, relatively, the same magnitude as the dataset average of $262 billion. Statistically, in any another context outside the scope covered in this document, the validity of the reported data would be heavily questioned due to the relative size of the dataset’s standard deviation. Validity and accuracy in reporting should, again, be paramount. Incorrectly reported values published without due diligence or values not corrected as new information becomes available is irresponsible. The sources of data used here have made no retractions or published updated information other than what is cited here.
The importance of journalism to society [39
] and particularly democracy [41
] is not only generally accepted, but often held up as the factor that makes democracy possible. The influence of mass media and journalism on policy in a number of arenas is also well established [44
]. Perhaps nowhere is this influence more important than when journalists inform the public about hazards; for hazards both commonplace and rare, the mass media is typically the primary source of knowledge for the public [52
]. This heavy responsibility falls upon the media because their published information will shape the public perception of the hazard and often drive the resultant policy. This is a non-trivial challenge as, in general, large-scale hazards such as nuclear power disasters are complex and scientific hazard reports are long and include great detail highlighting all aspects of risk or an incident. Thus, as Singer’s research indicates, this results in news stories that omit results, overgeneralize scientific findings, omit discussion of research design and methods, and in some cases overdramatize the scientific findings [52
]. This study investigates the potential impact of such journalistic failures by analyzing the effects of reporting on the inputs to calculating nuclear liability insurance subsidies and the resultant nuclear energy policy. As the results clearly show, mainstream media coverage of nuclear disaster costs is sorely lacking.
Some of these oversights can be partially explained by a general trend in reporting scientific information. Pulford’s [53
] research claims that most scientific articles commonly contain errors. Unfortunately, there is variance in the methods of publication utilized by the scientific community as compared to the media [54
]. Carsten and Illman further elaborate on the discourse between the two communities; scientists rely on objectivity and detail, while journalists write to both entertain and inform the public. Perhaps more important is the need for sufficient oversight of the generation of technical information. Maier’s cross-market analysis of 4800 news sources cited in 14 news articles identifies the newspapers’ accuracy and ultimate credibility [55
]. The news sources surveyed lack understanding, suffer deadline pressure, and perform insufficient research, leading to publishing errors in 25.9%, 18.9%, and 17.3%, respectively, of the articles analyzed [55
To help rectify this problem, there are several steps that can be taken. First, data should be accessible to the public (e.g., open access). With the advent of the Internet, journalists can now provide hyperlinks to direct sources, more in-depth explanations, actual studies, etc. from the electronic versions of their stories. Second, publication of scientific articles or news reports requires trustworthy, quantitatively accurate, and concise data. In cases of public subsidization of a particular technology/company are accompanied by acceptance of risks, the public must be quantitatively informed of those risks. If adequate resources (e.g., funding and support) are not available to the research entities (e.g., universities, NPOs, and government offices), the production of quality scientific data is difficult. Thus, thirdly, ensuring appropriate funding measures for scientific research into appropriate cost-benefit analysis of technologies that represent a hazard to the public is warranted and allows for legitimate policy discussion and reform.