Measuring lost recreational benefits in Fukushima due to harmful rumors using a Poisson-inverse Gaussian regression

The 2011 Great East Japan Earthquake hereafter, 'the Earthquake' and the accident involving radiation leakage at Tokyo Electric Power Company fs Fukushima Nuclear Power Plant No. 1 hereafter, 'NPP No. 1' brought fame to Fukushima Prefecture but not in a positive way. In general, economic damage caused by misinformation is defined as gdamage caused by groundless rumors, in particular, economic damage suffered by people or groups caused by improper news coverage, even though they have essentially nothing to do with an event or accident. h (Kojien sixth edition, 2008) This means that tourism in a given area will be affected by news coverage and misinformation that differ from the facts (for example, degradation in environmental quality at a recreation site), tourists will be deterred, and the economy at the site will be negatively affected. However, the degree to which such news coverage and misinformation affect people fs activities is largely dependent on those people fs state of mind. It is impossible to determine the exact number of visits that would have been made to the region in question had there been no such news coverage, no harmful rumors, and no environmental degradation. Thus, after the sensational news coverage about radiation at NPP No. 1, the inclusion of people in the survey sample who had never visited Fukushima Prefecture would have skewed the expected trip numbers and overestimated the monetary loss of tourism. This paper estimated the recreational benefits lost in Fukushima Prefecture due to rumor-driven economic damage from the NPP No. 1 radiation leakage accident in March 2011 to March 2014. Considering the hypothetical scenario in which a radiation leakage accident did not occur in Fukushima, we asked survey respondents how many times they would have visited the prefecture in this scenario and analyzed the responses using the Hypothetical Travel Cost Model. Since the survey participants were people who had actually visited the prefecture, we considered our data as pseudo on-site sampling. We thus expanded the Poisson-Invese Gaussian hereafter, 'PIG' regression model, which will improve the standard Poisson regression model in the analysis of count data with strong overdispersion, into a random effect model. In addition, to deal with the data collected through on-site sampling, we applied Shaw fs correction to the PIG random effect model and used it to estimate the demand function for the recreational trip. The estimation results showed that Fukushima Prefecture fs lost recreational benefits due to rumor-driven economic damage totaled approximately 2.85 trillion yen over the three years from the radiation leakage accident to March 2014.