Risk Perceptions and Amplification Effects over Time: Evaluating Fukushima Longitudinal Surveys

Abstract

Despite the vast body of literature addressing public risk perceptions of technological disasters such as nuclear accidents, longitudinal studies comparing pre- and post-disaster attitudes are scarce. This study examines (a) how perception factors such as concern, risk, political saliency, trust, and attitudes toward energy policies related to nuclear power changed over time and among different populations, and (b) if social amplification and ripple effects were present after the 2011 Fukushima Daiichi Nuclear Disaster in Japan. The article compares 2010 survey data collected before the nuclear accident with data collected in 2013 and 2016. The analysis shows a clear trend regarding concern over nuclear power. Levels of concern and political saliency increased in the aftermath of the nuclear disaster but declined in 2016. These findings indicate amplification effects between the pre-disaster survey in 2010 and the first follow-up survey in 2013. However, the 2016 data indicate the Fukushima accident only had a significant short-term impact on public risk perceptions. The data also indicate high levels of uncertainty regarding trust and future energy policies. The high levels of uncertainty present an opportunity for nuclear power supporters and opponents alike to shape Japan’s future energy policy through well-crafted communication and outreach programs.

1. Introduction

On 11 March 2011, a magnitude nine earthquake off Japan’s northeast coast triggered a massive tsunami that traveled up to 6 miles inland (Figure 1 and Figure 2). Around 19,000 people died, and more than 1 million buildings were entirely or partially destroyed [1]. It also led to the Fukushima Daiichi Nuclear Disaster, one of the country’s most devastating disasters with long-term impacts.

While the plant’s six reactors suffered no damage from the earthquake, flooding from the tsunami it generated resulted in power losses that crippled all external power supply sources. This power loss led to a shutdown of the cooling systems in Fukushima Daiichi towers 1, 2, and 3 [1]. A cascading effect resulted in reactor meltdowns and hydrogen explosions that released nuclear material into the atmosphere. The Japanese government declared a 20-km evacuation zone around the power plant. After discovering that radioactive material spread farther than anticipated, the government expanded the area.

The plant operator, Tokyo Electric Power Company (TEPCO), initially categorized the disaster as a level 4 event. This classification sparked widespread criticism by international organizations such as Greenpeace, which claimed the disaster was on par with the 1986 Chernobyl nuclear disaster ranked as a level 7 event [4]. The International Atomic Energy Agency (IAEA) supported Greenpeace’s evaluation [5]. According to the IAEA, extremely high levels of radioactive caesium were deposited northwest of the Fukushima Daiichi Power Plant, with contamination densities recorded between 1000 kBq/m² and 10,000 kBq/m². Despite widespread radioactive contamination decreasing, significant ongoing issues are a steady reminder of nuclear power risks for the Japanese people. Problems include the large-scale displacement of residents in the Tohoku region, land and water contamination through exposure to radiation and radioactive materials, and continuing challenges with melting fuel rods in severely damaged reactor buildings that are still highly radioactive.

Before the disaster, nuclear power supplied about one third of Japan’s energy. After the disaster, the Japanese government shut down all reactors as a precaution [6]. In 2013, the government introduced stricter safety standards that nuclear plants must meet before resuming operations to reduce the risk of damage from earthquakes and tsunamis. Simultaneously, the government increased nuclear energy back to about 20% of the country’s energy production to cut reliance on foreign fossil fuels [7]. By June 2019, only nine reactors and five power plants had met the new standards and resumed operations [8].

The Fukushima disaster and resulting reactions and policies raise interesting and important questions regarding risk, concerns, trust, and energy policies related to nuclear power. This article presents the results of a longitudinal study of survey data collected in Japan in 2010 (pre-disaster) and in 2013 and 2016 (post-disaster) on public perceptions in the directly impacted Tohoku region and the entire country. The purpose of this paper is two-fold. First, we explore the nature of public perceptions of nuclear power over time. We analyze the data to understand how and over what timespan the disaster impacted public attitudes, including opposition and support for nuclear power. Second, we attempt to identify if social amplification and ripple effects result in survey data changes over time. Specifically, the study addresses the following research questions:

• How have perceived levels of concern, political saliency of nuclear power, trust, and attitudes towards energy policies related to nuclear power changed over time and among different groups of the population?
• How do participants from the directly impacted Tohoku region compare with the rest of the country regarding perceived risks?

The paper also examines the role of social amplification of perceived risks and “ripple effects” as part of amplification. For example, one question asks if the Fukushima nuclear accident impacted the perception of the risks of global climate change. Despite the comprehensive body of literature addressing public risk perceptions of disasters such as nuclear accidents, longitudinal studies comparing pre- and post-disaster attitudes remain scarce [9,10]. The research presented here adds to the literature by examining how important perception factors, including concern, risk, political saliency, trust, and attitudes toward nuclear energy, changed among the Japanese population between 2010 and 2016. The analysis sought to understand if social amplification effects were present as well. The study also provides insights into whether nuclear power is still accepted as a viable climate change mitigation tool. It is projected that global temperatures will continue to increase if actions are not implemented to reduce total annual greenhouse gas (GHG) emissions [11], and nuclear power is hotly debated as part of the solution [12,13,14].

2. Literature Review

2.1. Trust

The role of trust is another significant influence on risk perception [15,16,17,18,19]. Trust is a dynamic process that takes place at the individual (cognitive, affective, and behavioral), institutional (policy, governance, organizations), and ideological (values and norms) levels [20,21,22,23,24]. Trust factors include perceived competence, objectivity, fairness, consistency, and goodwill. There are different ways trust is conceptualized. For instance, Xiao et al. [25] divided received trust into “goodwill trust” and “competence” in a Chinese resident study. Ryu et al. [26] classified trust into two groups: trust in government or management and trust in regulation. These studies concluded that various types of trust, directly and indirectly, influence perceived risk. Mushkatel et al. [27] argued that these studies did not address the role of trust in government.

Research revealed that various risk communication programs in Europe and the US had limited effectiveness [28]. The public’s trust in the communicator and those responsible for risk management determines its efficacy [29,30,31]. In most circumstances, new information is judged based on the source’s credibility [32]. If the source is not trusted, any message will likely be disregarded, no matter how well-intentioned. This statement is especially true for issues characterized by high uncertainty, such as nuclear power and technology. For these issues, trust plays a vital role in the success of risk communication and policy implementation.

There is also a positive correlation between trust and risk acceptance [33,34]. A study of risk perceptions toward China’s nuclear power by Huang et al. [34] established that the more people trusted the government to prevent nuclear accidents, the more they accepted nuclear power. Studies in OECD countries show that a lack of trust in government and power companies to manage and prevent accidents presents a challenge to nuclear energy [35]. Thus, trust is an underlying factor in the acceptance of nuclear power. The current study considers the relationships between trust, concern, and policy support in Japan over time. It also examines the role trust plays in amplifying risk perception.

2.2. Social Amplification and Ripple Effects

People dealing with uncertainties often over- or underestimate risks and threats [36]. Social amplification has been used to explain how people perceive, evaluate, and act on risks [37]. It posits that signals (e.g., images, signs, and symbols) communicate risks within social contexts [38]. Furthermore, multiple actors translate and interpret these signals in various ways and can amplify or decrease risk perception. As proposed in the Social Amplification of Risk Framework (SARF) by Kasperson et al. [38] and modified by several scholars [39], the social amplification framework aims to explain why risk events can lead to substantial public concern and social impacts. In the case of nuclear power, people tend to overestimate radiation exposure risks, resulting in significant social amplification effects and heightened concern over the technology.

Risk amplification can also have social and policy consequences beyond the area where a hazardous event occurs. Today, these consequences are often due to the media coverage following such an event. After the 2011 Fukushima nuclear disaster, social amplification of perceived risks of nuclear power influenced nuclear policies within Japan and internationally.

2.3. Risk Perceptions towards Nuclear Energy

2.3.1. International Response

A growing body of literature shows mixed responses to the Fukushima disaster. For example, there was a decrease in public support for nuclear power plant projects in the US, Taiwan, Italy, France, Sweden, and Spain, among other countries [40,41,42,43,44,45]. Ipsos Mori [46] captured public opinion on nuclear energy use across 24 countries shortly after the accident. More than 62% of respondents were against nuclear, including 26% who changed previously held views. This response is not new. Past accidents such as Three Mile Island (1979), Chernobyl (1986), and Tokai (1999) affected nuclear risk perceptions and attitudes [47,48,49,50].

Opposition to nuclear power led some countries to shut down their plants or cancel plans and pursue renewable energy. Belgium and Italy enacted legislation to phase out nuclear power in favor of more environmentally friendly energy sources [51,52]. Although more than 50% of Belgium’s electricity comes from nuclear, the country passed a law shortly after the Fukushima disaster to phase it out by 2025. Several groups criticized this action due to the lack of a well-defined replacement plan [52]. An Australian study showed a shift in public acceptance of nuclear energy as an alternative for tackling climate change [53]. Mah et al. [54] reported a significant shift in public risk perceptions in Hong Kong from data collected two months after the accident. However, a follow-up survey two years later showed that support for nuclear power had increased.

Other countries continued public policy support for nuclear energy after the Fukushima accident. In Indonesia, Biddinika et al. [55] found that the accident barely impacted its policy toward nuclear power. A United Kingdom study showing public perceptions by Poortinga et al. [56] showed that nuclear power acceptance did not decline. France and China stayed committed to their nuclear plans but developed more stringent safety measures and regulations [57]. These studies provide important insights into the social amplification and ripple effects of nuclear energy outside areas directly impacted by the Fukushima disaster.

2.3.2. Japanese Response

Studies have also explored Japanese perceptions about nuclear power and attitudes towards decision-makers since the 2011 disaster [58,59]. A 2012 poll showed a significant loss of trust in nuclear safety and high dissatisfaction with the government and TEPCO [60]. Using an ordered probit model, Arikawa et al. [58] found that socioeconomic status, particularly gender and education, influenced the acceptance of Japan’s continued use of nuclear power. For instance, highly educated males demonstrated unwavering support.

Changes in public opinion on Japan’s energy policies have surfaced in several studies, despite the traditional restraint from challenging the government’s decision-making authority [58]. Before the 1990s, public opposition to nuclear energy mostly came from those close to power plants [61]. The most common reason was health concerns. These were short-lived as people got used to living near a plant. As more nuclear reactors were built, risk became more of a concern. A series of accidents during the 1990s also suggested a lack of proper management and monitoring of nuclear plants. These accidents led to a loss of public trust and confidence in the Japanese government’s ability to manage risk [62], and criticism of the management of their nuclear program mounted as well.

Other studies have linked public perceptions and attitudes towards nuclear power with topics such as energy security or climate change attitudes, testing how the various perceptions influence each other [56,63,64,65,66]. For example, Corner et al. [63] investigated public perceptions of nuclear power in Great Britain as an alternative to climate change. The study shows a positive relationship between the level of concern regarding climate change and the acceptance of nuclear power as a sustainable energy source and mitigation strategy.

Research has also sought to understand the impacts of perceptional factors on behavior and policy support [56,64]. In a study of community attitudes, Venables et al. [67] discovered that those respondents who opposed nuclear power perceived significant threat levels. Visschers et al. [68] also found a negative relationship between risk perception and nuclear energy acceptance. However, this was not as strong as the positive relationship between perceiving nuclear power as a reliable energy source and as a climate change mitigation tool. DeGroot et al. [69] show that perceived risk is also related to the willingness to support or oppose nuclear energy actively. Other studies link perceptions to concerns about accident risks, waste management, and terrorist attacks [70,71,72].

One of the few longitudinal studies, Visschers and Siegrist [9], found only a moderate impact of the Fukushima accident on the long-term acceptance of nuclear power. Therefore, they suggest that people have relatively stable attitudes towards this energy source. The study presented in this article adds to this body of literature by looking at longitudinal data from three surveys conducted in 2010, 2013, and 2016.

3. Methodology

3.1. Survey Design

The data analyzed and presented in this study are part of two larger survey instruments. The first survey instrument was used in the 2010 data collection. The survey instrument was divided into thematic sections addressing topics such as the political saliency of various socio-economic issues, including climate change and energy policy. Other topics included perceived levels of concern and threat towards environmental and technological risks, level of trust towards various risk communicators and managers, levels of support for different risk mitigation and adaption strategies, and willingness to commit to behavioral changes. Following the disaster in March 2011, the survey instrument was updated to include additional questions specifically acknowledging the 2011 Tsunami and the resulting Fukushima Daiichi Nuclear Disaster. This updated survey instrument was used during the data collections in 2013 and 2016.

The data discussed focuses on the following public perceptions, attitudes, and phenomena:

(a)

Concern and political saliency regarding environmental protection, nuclear safety, and renewable energy;

(b)

Trust towards the government and TEPCO;

(c)

Attitudes towards various energy policies and strategies;

(d)

Ripple effects from the nuclear accident and level of knowledge impacting public perceptions of climate change.

The redesigned survey questions were similar to previous studies focusing on natural and technological risks [73,74,75,76,77]. Most questions were closed-ended and multiple-choice, making it easy to code and analyze statistically [78]. Response choices used Likert-type scaling to prevent bias with an equal number of favorable and unfavorable answer categories. The Likert scales used in the relevant survey questions for this study were 4- to 5-point scales with answers ranging from:

(a)

4-point-Scales

o

“unimportant” to “very important”

o

“not informed” to “very informed”

(b)

5-point scales

o

“strongly distrust” to “undecided” to “strongly trust”

o

“strongly disagree” to “undecided” to “strongly agree”

o

“strongly reduced” to “undecided” to “strongly increase”

o

“not at all concerned” to “somewhat concerned” to “very concerned”

The survey instrument was tested and reviewed by experienced researchers to ensure its validity. Data were analyzed using frequency distributions, descriptive statistics, and chi-square and correlation tests to establish significant relationships between different groups within the sample. The tests showed a significant relationship between the survey year, participant location (Tohoku region versus the rest of the country), and perception factors.

3.2. Data Collection & Data Characteristics

Survey Sampling International provided the internet panels, hosted the survey, and oversaw data collection in 2010 (pre-Fukushima accident), 2013, and 2016 (post-Fukushima accident). We oversampled the Tohoku region for the 2013 and 2016 surveys to ensure a representative sample. A cross-section of those 18 years and above included all gender, income groups, education levels, and different regions. The total sample size was 3402 households, with 829 surveyed in 2010, 1102 in 2013, and 1471 in 2016 (

Table 1. Sample Sizes.

Year of Survey	Population Size		Survey Sample		Confidence Level	Margin of Error
	Region	Total	N	% of Total
2010	Tohoku Region	9,335,636	44	0.001%	95%	14.8%
	Remaining Japan	118,721,364	785	0.001%	95%	3.5%
2013	Tohoku Region	9,335,636	452	0.005%	95%	4.61%
	Remaining Japan	118,721,364	650	0.001%	95%	3.85%
2016	Tohoku Region	8,983,000	500	0.006%	95%	3.68%
	Remaining Japan	126,933,000	971	0.001%	95%	3.09%

).

Internet panel surveying has several advantages over other survey methods (e.g., telephone, mail, or in-person surveys). First, it allows coverage of a wide geographical area with relatively low costs, ensures a demographically representative sample for the entire country, and establishes consistency and standardized collection to gather quantifiable, empirical comparative data. Notably, it enables participants to answer questions at their convenience, increasing response rates [79]. Furthermore, participants remain anonymous throughout the entire process. Since the database assigns an ID number to each person without their name or address, it is impossible to link answers to a specific person. Anonymity is paramount when surveys address personal feelings and behavior.

For most cases, the sample size is large enough to generalize results with a 95% confidence level and ±4% margin of error. There is one exception. The Tohoku region was not of specific interest since the 2010 survey data was collected before the nuclear power plant disaster. Instead, the 2010 survey aimed only at a representative sample of the entire country. As a result, the margin of error with a 95% confidence level for the Tohoku sample in 2010 is very high at 14.8%, while the 2013 sample of the Tohoku region is slightly above the 4% threshold at 4.61%.

4. Results

4.1. Concern and Political Saliency

4.1.1. Concern

The survey instrument for all three years (2010, 2013, and 2016) asked how concerned survey participants were about the environment to establish a baseline (

Table 2. Levels of concern about the environment.

About the Environment
Year	Sample	1 Not at All Concerned	2 Slightly Concerned	3 Somewhat Concerned	4 Concerned	5 Very Concerned	Mean	Std. Dev.
2010	Tohoku Region	2.3%	34.1%	22.7%	20.5%	20.5%	3.23	1.199
	Remaining Japan	2.3%	37.6%	28.2%	19.7%	12.7%	3.02	1.075
2013	Tohoku Region	0.4%	10.6%	40.3%	32.7%	15.9%	3.53	0.899
	Remaining Japan	1.4%	7.1%	36.0%	40.0%	15.5%	3.61	0.880
2016	Tohoku Region	2.2%	10.8%	45.4%	26.6%	15.0%	3.41	0.945
	Remaining Japan	1.5%	10.5%	41.2%	30.7%	16.1%	3.49	0.935

).

The data show that concern levels changed over time. Specifically, levels increased from 2010 to 2013 following the disaster but decreased in 2016. However, people were still more concerned about the environment five years after the accident than in 2010. Similar findings were reported in studies among British populations [56], where concern over the environment and Global Climate Change (GCC) increased following the Fukushima accident.

The 2013 and 2016 surveys also included a question on the level of concern about nuclear technology safety not included in the 2010 survey (

Table 3. Levels of concern about nuclear technology in Japan.

About the Safety of Nuclear Technology in Japan
Year	Sample	1 Not at All Concerned	2 Slightly Concerned	3 Somewhat Concerned	4 Concerned	5 Very Concerned	Mean	Std. Dev.
2013	Tohoku Region	0.4%	6.2%	25.0%	27.4%	40.9%	4.02	0.974
	Remaining Japan	1.4%	9.4%	22.8%	32.8%	33.7%	3.88	1.026
2016	Tohoku Region	1.6%	12.0%	29.2%	26.2%	31.0%	3.73	1.075
	Remaining Japan	2.8%	12.7%	29.7%	25.3%	29.6%	3.66	1.112

).

In 2013, 68.3% of Tohoku region residents indicated they were ‘concerned’ or ‘very concerned,’ whereas only 66.5% of other Japanese residents shared that sentiment. Interestingly, the levels of concern dropped from 2013 to 2016, indicating that acceptance bounced back to near pre-disaster levels in a relatively short period. In the Tohoku region, scores decreased by 11.1% to 57.2% ‘concerned’ or ‘very concerned’. For other residents, scores dropped by 11.6% to 54.9%. Similar trends were recorded among populations in Europe, China, and Australia [34,46,51,53,56], where public skepticism and increased concern over the safety of nuclear technology rose right after the Fukushima accident.

4.1.2. Political Saliency

Surveys measured political saliency by asking participants to indicate how important it is for government to act on nine problem areas in 2010 and ten in the 2013 and 2016 surveys. All studies included ‘moving towards a renewable energy policy’ (Figure 3). 2013 and 2016 surveys also included ‘improving the safety for nuclear power’ (Figure 4).

In 2010, 36.4% of Tohoku survey participants and 35.9% of other Japanese respondents ranked renewable energy policy as very important. These numbers rose significantly following the 2011 events, with 58.6% of Tohoku residents and 53.1% of other residents identifying it as ‘very important’ in 2013. By 2016, renewable energy had moved back down the political agenda, with mean scores at or near 2010 levels. Studies in Germany [43], Indonesia [55], and internationally [46] also show a significant increase in the public’s support for renewable energy in the aftermath of the nuclear accident.

The political saliency of improving nuclear power safety follows a similar narrative. Whereas improving nuclear safety was perceived as ‘very important’ by 56.9% of all survey participants of the Tohoku region and by 52.2% of all other survey respondents in 2013, only 46.4% and 46.8%, respectively, ranked it as ‘very important’ in 2016.

While the study did not specifically ask participants to provide a reason for lower levels of concern, it did seek to determine if participants’ level of trust in the government and TEPCO to manage risk had changed significantly.

4.2. Trust in Government and TEPCO

Trust is a critical perceptual dimension that influences policy and organizational success. Therefore, public trust in organizations responsible for managing risks that carry high uncertainties (e.g., nuclear power) is vital to generate support and increase their likelihood of success. The 2013 and 2016 surveys asked about the public’s level of trust in (a) TEPCO and (b) the Japanese government in managing the safety of nuclear technology. Figure 5 and Figure 6 show trust levels improving since the 2011 accident. However, the results show that a substantial portion of Japan’s population remains skeptical or uncertain of TEPCO’s and the government’s capacity to manage nuclear power risks.

In 2016, 90.2% of respondents in the Tohoku region were undecided (34.6%) or distrustful (55.6%) of TEPCO in managing the safety of nuclear technology. While the remaining country respondents expressed increased trust in TEPCO (14.5%), over 85% who remained distrustful (54.4%) or undecided (31.1%). The levels of trust in the government were slightly higher than those for TEPCO, with a significant percentage still somewhat distrustful (30.1%) and undecided (32.3%) in 2016.

The surveys also asked respondents to what level they agreed or disagreed with the statement “the Japanese government has done enough to prevent another nuclear accident”. The responses show low levels of trust. In 2013 and 2016, over 50% of respondents ‘strongly disagreed’ or ‘disagreed’ that the government has done enough. Around 30% were ‘undecided’ and less than 18% ‘agreed’ or ‘strongly agreed’.

The last question addressing trust asked participants how the government’s approach to the 2011 nuclear disaster influenced their perception of nuclear power plant risks. The 2013 results show high levels of uncertainty, with uncertainty beginning to resolve by 2016. For instance, 31.1% of survey participants in the Tohoku region and 26.6% of respondents from other parts of the country stated that the government’s handling of the nuclear crisis strongly increased their perceived level of risk of nuclear power plants in 2013. By 2016, these numbers declined to 17.2% and 16.7%, respectively.

4.3. Energy Policies and Strategies

4.3.1. Renewable Energy

The survey also examined public support for two groups of energy policies over time: renewable energy, in general, and nuclear energy, specifically. For renewable energy, the surveys asked participants how much they support or oppose (1) the government providing subsidies to the private sector to invest in alternative energy and (2) a policy requiring electric utility companies to produce at least 20% of their electricity from renewable energy sources by 2020.

The responses to the subsidies concept varied but were generally positive. All samples had a similar level of ‘moderate’ to ‘strong’ support in 2010, with Tohoku slightly higher than the remainder of the country. The ‘Strongly support’ data show a spike in 2013 but dropped closer to pre-disaster levels in 2016. The mean scores indicate that overall support has declined over time.

The data regarding requiring the increased use of renewable energy indicate a less homogenous picture between Tohoku and the rest of the country. The Tohoku data show strong public support rising from 4.5% (2010) to 16.8% (2016). Support in the rest of the country spiked in 2013, with strong support increasing from 10.8% to 16.8% but declining to 12.6% in 2016. The mean scores and combined ‘moderately support’ and ‘strongly support’ categories for Japan excluding the Tohoku region demonstrate that overall support for the policy has declined.

4.3.2. Nuclear Energy

The second group of policies focused on nuclear energy as (1) a tool to mitigate climate change and (2) an essential cornerstone of Japan’s energy portfolio. The results, shown in Figure 7 and Figure 8, suggest that the Japanese population is still reluctant to embrace nuclear energy.

Despite slowly increasing support across Japan, most people (over 75%) did not favor nuclear power in 2013. Among the survey respondents in the Tohoku region and the remaining country, over 50% ‘strongly disagreed’ or ‘disagreed’ with the notion that it would be important for the country to continue to use nuclear power as a major energy source (Figure 7). Moreover, the idea of introducing more power plants to reduce fossil fuel needs and reduce GCC also was viewed very skeptically by survey respondents, especially since the Fukushima disaster. As shown in Figure 8 and prior to the Fukushima disaster, 11.4% of the Tohoku region ‘strongly disagreed’ with the notion of introducing more nuclear power plants. In 2013, just over two years after the disaster, 29% strongly opposed the idea. By 2016, the strong opposition had declined by 8.6 percentage points with 20.4% of the survey respondents from the Tohoku region strongly disagreeing with the idea of building more nuclear power plants. A similar trend can be observed among the remaining surveyed population outside of the Tohoku region with 8% strongly disagreeing in 2010, 24.2% strongly disagreeing in 2013, and 17.9% strongly disagreeing in 2016.

For all population groups sampled, the support for introducing more nuclear power to mitigate climate change had not reached pre-disaster levels in 2016. This slow increase suggests that the public still distrusts nuclear technology.

Like trust, public attitudes toward energy policies show high levels of uncertainty and undecidedness. Between 33% and 50% of the survey respondents had no clear opinion on the proposed energy policies and strategies. With energy technologies and supply being hotly debated, the people who are undecided today could determine what Japan’s energy policy will look like in the future.

4.4. Ripple Effects

4.4.1. Global Climate Change (GCC)

One characteristic of stigma-related effects is how highly visible events with significant impacts can change social consciousness, behavior, and perceptions on issues unrelated to the event. For example, distrust of the government and industry to manage the nuclear energy risks can cause a ripple effect and lead to a lack of trust in their ability to manage risks associated with other technologies. The 2013 and 2016 surveys asked about GCC and changes in concern levels and attitudes after Fukushima to test for social stigmatization and ripple effects. First, the survey asked respondents “to what degree did the government’s approach to the Fukushima accident influence GCC risk perceptions” (

Table 4. Government’s impact of handling the nuclear accident on global climate change risk perceptions.

Year	Sample	1 Strongly Reduced	2 Reduced	3 Undecided	4 Increased	5 Strongly Increased	Mean	Std. Dev.
2013	Tohoku Region	0.0%	0.9%	49.1%	33.8%	16.2%	3.65	0.754
	Remaining Japan	0.2%	1.2%	49.2%	35.4%	14.0%	3.62	0.742
2016	Tohoku Region	0.8%	3.6%	64.2%	21.2%	10.2%	3.36	0.746
	Remaining Japan	0.8%	3.7%	65.4%	19.5%	10.6%	3.35	0.752

).

In 2013, 50% of all survey respondents in the Tohoku region and 49.4% in remaining Japan stated that the government’s handling of the Fukushima accident ‘increased’ or ‘seriously increased’ perceived GCC risks. By 2016, the influence on GCC risk perceptions had declined to 31.4% and 30.1%, respectively. GCC risk perceptions and social amplification declined by 19.6 percentage points in a three-year span in the Tohoku region and by 19.3 percentage points in the remaining country. This significant drop points to how ripple effects diminish over time. Despite this decline, the data support the fact that stigma-related effects ripple to other areas of concern.

4.4.2. Environmental Change

The survey also sought to understand the impacts on the perceived level of consequences of various environmental changes, including GCC, over the next 20 years. Participants rated eight potential scenarios ranging from deterioration of the ozone layer to increasing drought frequency to worsening air and water pollution. In 2010, 82.6% of all respondents stated that they expect to experience ‘moderate’ or ‘serious environmental consequences’ over the next 20 years. Responses increased to 93.3% in 2013 then decreased to 90.5% in 2016. Only 17.5% of participants stated that they expect ‘no’ to ‘slight negative consequences’ from potential environmental changes before the Fukushima disaster. These responses decreased significantly in 2013 (6.8%) and 2016 (9.5%).

In the Tohoku region, 79.2% expected ‘moderate’ to ‘serious negative consequences’ in 2010. Responses increased to 93.7% in 2013 and 89.8% in 2016. These increases suggest that first-hand experience of a significant event with long-term impacts can have a psychological effect and increase perceived risks in other areas. However, more research is needed to establish causation between these types of events and higher environmental risk perceptions overall.

4.5. Knowledge and Awareness

A noxious event can impact people when they are not aware of or feel inadequately informed about the inherent risks. This can increase levels of anxiety and uncertainty especially when health and well-being concerns are significant [17]. The 2013 and 2016 surveys captured self-reported knowledge data regarding (a) the current safety situation at the Fukushima power plant, (b) the actions taken to resolve the issues at the plant, and (c) the future environmental and health hazards resulting from the accident (

Table 5. Self-reported knowledge on different aspects of the Fukushima Daiichi Nuclear Disaster.

(a) Current Safety Situation at the Fukushima Power Plant
Year	Sample	1 Not Informed	2 Somewhat Informed	3 Informed	4 Very Informed	Mean	Std. Dev.
2013	Tohoku Region	21.5%	47.1%	22.3%	9.1%	2.19	0.875
	Remaining Japan	22.6%	48.2%	22.8%	6.5%	2.13	0.834
2016	Tohoku Region	25.2%	36.2%	25.0%	13.6%	2.27	0.987
	Remaining Japan	28.5%	39.8%	23.5%	8.2%	2.11	0.915
(b) Actions to resolve the Fukushima Power Plant Issues
Year	Sample	1 Not Informed	2 Somewhat Informed	3 Informed	4 Very Informed	Mean	Std. Dev.
2013	Tohoku Region	28.3%	48.9%	17.5%	5.3%	2.00	0.820
	Remaining Japan	28.8%	49.2%	16.6%	5.4%	1.99	0.819
2016	Tohoku Region	35.2%	37.2%	20.0%	7.6%	2.00	0.926
	Remaining Japan	37.7%	39.6%	17.0%	5.7%	1.91	0.875
(c) Current and Future Environmental & Health Hazards Resulting from the Fukushima Power Plant Accident
Year	Sample	1 Not Informed	2 Somewhat Informed	3 Informed	4 Very Informed	Mean	Std. Dev.
2013	Tohoku Region	18.4%	49.3%	23.2%	9.1%	2.23	0.853
	Remaining Japan	21.7%	50.8%	21.8%	5.7%	2.12	0.807
2016	Tohoku Region	21.6%	40.2%	26.2%	12.0%	2.29	0.937
	Remaining Japan	23.4%	47.0%	21.2%	8.4%	2.15	0.873

).

Combining both the Tohoku region sample and the survey participants from the remaining country, the data show that 70.2% of the nation felt ‘not informed’ (23.0%) or only ‘somewhat informed’ (47.2%) in 2013 regarding the current safety situation at the Fukushima Power Plant. The percentage was similar in 2016 (68.1%) with slightly more people feeling ‘not informed’ (28.7%) and fewer people feeling ‘somewhat informed’ (39.4%). This lack of knowledge was the same throughout Japan, including the directly impacted Tohoku region where 68.6% felt ‘not informed’ or only ‘somewhat informed’ in 2013 and 61.4% in 2016.

The survey also asked if respondents felt sufficiently informed about the actions taken by the government and TEPCO to resolve issues following the disaster. In 2013, 78.4% of all respondents felt either ‘not informed’ (29.3%) or ‘only somewhat informed’ (49.1%). Data from 2016 showed similar overall results as well as increasing uncertainty. For instance, 29.3% stated that they did not feel informed at all in 2013 and 38.3% voiced the same sentiment in 2016. The data suggests that proximity to the Fukushima Daiichi nuclear power plant does not meaningfully impact the public’s self-reported knowledge regarding past, present, or future planned actions to resolve the issues within the Fukushima Daiichi power plant.

The final question captured perceptions about current and future environmental and health hazards resulting from the accident. Over 60% of all participants surveyed in both 2013 and 2016, expressed that they were ‘not informed’ or only ‘somewhat informed’ about hazards. Overall, these results suggest that people felt that they lacked important knowledge or received inadequate information about the nuclear disaster. These findings are supported by the June 2011 Ipsos global survey [46]. The survey shows that among the 24 participating countries, including Japan, only between 1% (South Korea) and 34% (Turkey) fully agreed with the sentiment that Japanese officials and their government leaders communicated the nature, impact, and mishaps of the Fukushima nuclear disaster appropriately. In Japan, only 4% fully agreed with that statement.

5. Discussion & Conclusions

5.1. Concern and Political Saliency

The data show a clear trend regarding concern over nuclear power, the environment in general, and the perceived political saliency of increasing the amount of renewable energy or improving the safety of nuclear power. Levels of concern and political saliency increased in the aftermath of the nuclear disaster but declined in 2016. These findings indicate amplification effects between the pre-disaster survey in 2010 and the first follow-up survey in 2013. However, the 2016 data indicate the Fukushima accident only had a significant short-term impact on public risk perceptions. Levels of concerns regarding nuclear technology declined with acceptance bouncing back to pre-disaster levels by 2016.

The changes are in line with recent voting patterns in Japan, where the pro-nuclear candidate Shinzō Abe was elected as prime minister in 2012 then re-elected in 2014 and 2017. For instance, the data suggest that the renewed goal to increase nuclear energy to 20% of the country’s energy production might have slightly less public opposition than expected after such a devastating disaster with long-term impacts (54.7% moderately or strongly support the policy). The data indicate, however, that the slow recovery of trust in TEPCO (54.0% ‘strongly’ or ‘somewhat’ distrust the company in 2016) and the Japanese government (46.3% ‘strongly’ or ‘somewhat’ distrust the company in 2016) may present a substantial roadblock.

5.2. Trust

The analysis reveals that distrust levels spiked in 2013 but declined in 2016, again pointing to short-term social amplification effects. It is essential to point out that compared to the levels of concern, levels of distrust declined more slowly. Furthermore, the questions asking whether the government is doing enough to prevent a similar nuclear accident in 2016 show that over 50% disagreed. These results suggest that distrust in the government’s nuclear risk management capabilities still exists five years after the accident.

5.3. Energy Policies

While perception changes in concern, political saliency, and trust show clear signs of social amplification effects, the picture is less clear regarding changing attitudes towards energy policies over time and between participants from the Tohoku region and the rest of the country. Public perception and support vary across samples and time. The data convey continued support for policies that reduce reliance on nuclear power in the Tohoku region, compared with only a 2013 spike in support in the rest of the country and declining support in 2016. These differences suggest a more permanent impact on attitudes to policy within the Tohoku region in contrast to a temporary effect for the rest of the country.

In terms of the pro-nuclear policies explored, the 2013 and 2016 data indicate an increase in public support over time of varying degrees between different population samples. Strong support was declining within the Tohoku region, whereas the percentage of people voicing strong support for pro-nuclear policies beyond Tohoku was increasing. Notably, nuclear power support in 2016 had not reached pre-disaster levels.

5.4. Stigma and Ripple Effects

The presented research shows that ripple effects for the Fukushima disaster are not limited to public attitudes and policy changes directly related to nuclear technology. Instead, the Japanese public projected their disappointment and lack of trust in their government as a viable risk manager for climate change threats. The Fukushima disaster’s perceived mishandling led to a spike in perceived climate change risk. Six years after the accident, survey data show GCC risk perceptions declining again. Significantly fewer people reported that the government’s handling of the nuclear accident ‘increased’ or ‘seriously increased’ their perceived climate change risks. This drop points to how social amplification and ripple effects diminish over time. Despite this decline, the data demonstrate that ripple effects around nuclear risk transferred to other areas of concern.

5.5. Study Limitations and Future Research

The study presented here had several limitations due to the reliance on anonymous internet panels for data collection. First, it is unknown what kind of technical information respondents had access to before and just after the nuclear accident and tsunami. This level of knowledge most certainly has an impact on the survey results. Second, it is possible that a significant amount of the voiced uncertainty among the survey participants comes from the fact that the nuclear disaster was in large part the result of a geological phenomenon, the earthquake, which cannot be predicted. Third, since all survey participants are anonymous it is highly unlikely that any of surveyed population participated in all three surveys between 2010 and 2016. This limits the statistical tools available for analysis.

Despite these limitations, the survey demonstrates that the Fukushima Daiichi Nuclear Disaster impacted public perceptions and policy in Japan, and reenergized the debate around the risks of nuclear power and the need for alternative and renewable energy sources in many other countries. It would be very interesting to see how public perceptions have changed in this context in countries that announced changes following the 2011 event. Therefore, this study should function as a benchmark for follow-up research adding more countries and longitudinal data to the database as well as other ethnographic research designs.