A Case from the Great East Japan Earthquake and Tsuna

Home , 1995 in Japan, Kesennuma, Reconstruction Agency

Progress in Disaster Science 9 (2021) 100141

Contents lists available at ScienceDirect

Progress in Disaster Science

journal homepage: www.elsevier.com/locate/pdisas

Regular Article Exploratory case study for neighborhood participation in recovery process: A case from the great East Japan earthquake and tsunami in Kesennuma, Japan ⁎ Kensuke Otsuyama a, , Rajib Shaw b a Research Center for Advanced Science and Technology, The University of Tokyo, 4-6-1 Komaba, Meguro-ku, Tokyo 153-8904, Japan b Graduate School of Media and Governance, Keio University, 5322 Endo, Fujisawa-shi, Kanagawa 252-0882, Japan

ARTICLE INFO ABSTRACT

Article history: About ten years have passed since the Great East Japan Earthquake and Tsunami. In the context of Japan, neighbor- Received 23 July 2019 hood voluntary civic associations (NVCA) play signiﬁcant roles in communicating, sharing, and building consensus Received in revised form 24 October 2020 at grassroots level in Japan, though it is unclear what are the elements for the successful engagement of disaster sur- Accepted 11 November 2020 vivors at the neighborhood level. This study examined whether community activities with the direct participation Available online 20 January 2021 by disaster survivors enhance DRR activities, or external stakeholders strengthen those activities. Through the ﬁeld survey in Kesennuma City, asymmetric results imply that the NVCA activities diverged from sociohistorical Keywords: Disaster recovery background. Great East Japan Earthquake and Tsunami Participation multi stakeholders' involvement Civic association

1. Introduction approach with the government reconstruction efforts. Here we will use ‘direct participation’ as participating by disaster survivors in the decision- Once in a millennium event, the Great East Japan Earthquake and Tsu- making process at grassroots level. Iiduka [19]alsopointedoutthatanin- nami (hereafter referred to as GEJET), occurred on March 11, 2011. The termediary organization also functions to connect and coordinate neighbor- death toll was 19,729 and still missing 2559 bodies as of March 2020, hood organizations with external resources such as NPO/NGOs or research and more than 120 thousand houses were destroyed [34]. Although institutions, while Inagaki [20] cautioned that intermediation would fail if about 10 years have passed since the event and the recovery process has the external group underestimated the self-motivation of disaster survivors somewhat progressed, roughly 1850 people are still living in temporary and neighborhood groups. Then, what are the elements for the successful houses as of November, 2020 [34]. Through the several devastating disas- engagement of disaster survivors for rebuilding the neighborhood? This is ters over the decades since the Hanshin-Awaji Earthquake (1995) in Japan, a central research question in this study. the natural hazards were beyond the government ability to minimize the Authors conducted field survey through the questionnaire and non- risks particularly after the GEJET in 2011. The national legislation for disas- structured interview, focusing on neighborhood association which rooted ter risk reduction (hereafter referred to as DRR), the Disaster Countermea- local history and indigenous cultural boundaries in Japan. The selected re- sures Basic Act was amended in 2013 to promote community based DRR search field was Kesennuma City in Japan. Kesennuma suffered from the activities at neighborhood level [9]. third highest mortality in the GEJET impacted municipalities in which a Past literature reported that top-down recovery plans sometimes made couple of communities were destroyed more than 60% of the houses, trouble such as forced migration due to the readjustment of zoning, prohibit while the subsequent processes are diverse. The interview based on the habitation, and unqualified housing due to the insensitivity to the local en- questionnaire survey revealed the asymmetric situation at the post- vironment [2,11,21,33]. Ganapati and Ganapati [16] also critically pointed disaster phase because Kesennuma has experienced catastrophes not only out that the narrow definition of ‘public’ caused exclusion from participa- the GEJET but two more huge tsunamis for 120 years. Disaster survivors tion by local communities and disaster victims in the recovery housing pro- have reconstructed the city again and again. In this exploratory case jects. In other words, direct participation by disaster survivors (or victims) study, the authors aim to understand how the previous lessons were trans- to the decision-making process rather than participation by selected stake- formed into lessons or decayed over the century. To answer this, the rela- holders, is the key for rebuilding the neighborhood as a complementary tionship of citizens' participation intersecting with indigenous culture and

⁎ Corresponding author. E-mail addresses: [email protected] (K. Otsuyama), [email protected] (R. Shaw).

http://dx.doi.org/10.1016/j.pdisas.2021.100141 2590-0617/© 2021 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/ 4.0/). K. Otsuyama, R. Shaw Progress in Disaster Science 9 (2021) 100141 external networks were examined as potential variables of questionnaire 2.3. Selected variables and research framework survey, and then non-structured interviews were administered to two se- verely impacted communities in the city. As an exploratory case study, this paper picked up two perspectives: in- This exploratory case study might contribute to identify the potential ternal resources for voluntary neighborhood association activities and ex- variables which promote community driven DRR activities at neighbor- ternal stakeholders. This paper examined how these two perspectives hood level, but the results limit its generalization to all types and place of were interrelated to DRR activities at NVCA level and District Association disasters due to the deeply rooted local context. Therefore, the key take- level in a city. away of this study would inform other researchers for the customized var- Shaw et al. [35] mentioned that indigenous culture or cultural succes- iables based on the local context. sion strengthens the neighborhood in bonding through intergenerational interactions. As local festivals or cultural events are managed annually, management members have shared the responsibility, leadership, and ad- 2. Research framework ministration tasks. Wilson [47] also stated: “Social memory leads to a learning and adjustment phase based on past experience”; local lessons need to 2.1. Community participation and involvement other stakeholders translate into activities. For example, the level of participation is based on whether they have initiatives for disaster risk reduction activities such as The forms of participation are broadly varied [3,17]: for example, par- evacuation drills and updating evacuation maps. As ( [15], p. 259) indi- ticipating as part of the labor force for housing recovery reconstruction, in- cated, “memory is the accumulated experience and history of the system, volving only stakeholders for house builders and donors excluding citizens and it provides content and sources for renewal, recombination, innova- [16], or joining in the decision-making process in post-disaster phases [13]. tion, novelty and self-organization following disturbance.” In the context One of the issues on the intersection of community participation is its selec- of Japan particularly the GEJET impacted area, local traditional performing tion of stakeholders concerning the size and inclusive involvement to and festivals contribute to communicate over the generations and to suc- achieve wider, but not competitive, participation in pre- and post-disaster ceed its tradition at the neighborhood level [46]. Also, tea parity and health planning [5,8,10]. Although stakeholders are represented from various sec- exercise and neighborhood level function as communication tools for com- tors including local industries, non-profit organizations, and elected offi- munity bonding, therefore these activities were used for soft interventions cials as local decision-makers, the gap between disaster survivors and to enhance communication in disaster temporary housing in the post- represented stakeholders is frequently reported [7,21,28]. Case studies in disaster phase [27]. It is unknown that these neighborhood activities di- Kobe Earthquake (1995) vividly demonstrated that direct participation by rectly promote DRR activities at NVCA or District level, though this paper disaster survivors and intermediate role of external planning professional used them as exploratory case study. as external human resources were successful completed its consensus for Literature on social network and disaster recovery were reviewed with a the planning at the neighborhood level [2,31]. In other words, matching matrix by Jones and Faas [22]. The matrix consists of egocentric networks of internal and external resources for neighborhood recovery planning a and sociocentric networks which are divided by individuals and organiza- key approach. tions. Egocentric network defined as “ties in each individual's or organization's own network, while sociocentric network as “an identifiable 2.2. Neighborhood voluntary civic association in Japan group or interaction sphere” [22]. As an egocentric network study, Steven- son and Conradson [39] found that preexisting relationships among 32 or- Neighborhood associations have played crucial roles in mobilizing local ganizations including social services, manufacturing, retails, and so on, resources, networking, and sociopolitical activities [4,6]. Lee and Franser performed for communication in urgent situations during the earthquake [29] also noted that civic association is the highest type of participation in Canterbury. On the other hand, for sociocentric network study, Harris in voluntary associations in Japan and found that more active voluntary as- and Doerfel [18] identified organizational centrality through the local sociations had more disaster survivors. Jichikai, or neighborhood voluntary stakeholder network analysis in the Hurricane Sandy post-disaster phase di- civic associations (hereafter referred to as NVCA) plays a significant role for vided four-time phases in the first three months since the event. They found communicating, building consensus, and recreation among residents at that network transition from federal organizations at the emergency phase neighborhood level in Japan. Ministry of Internal Affairs and Communica- to local non-profit organizations and case management [18]. It makes sense tions [30]defined NVCA as “an organization consisting of residents who that shrinking and customizing the network into the local context as time have addresses in a certain boundary, and it plays various roles as coordina- passes. Oikawa [32] argued that networks played a critical role in emer- tion, cooperation, communication, and maintenance of community facili- gency response and the recovery phase especially non-governmental orga- ties.” There are 298,700 NVCA in Japan as of 2013 [30]. Furthermore, nizations (NGOs), nonprofit organizations (NPOs), or research institutions Jichi-Rengo, an uppermost organization in a district which consists of contributed to impacted locals at grassroots level in the GEJET. In the con- NVCA leaders (hereafter referred to as District Association), plays a man- text of Japan, aging society, the Local Social Welfare Council (LSWC) works agement role in building consensus and sharing official information from for elderly care but one of the common interests with NVCA is disaster man- municipal administration in a district. The District Association functions agement in the local communities [42]. Hence, networking with LSWC is a as a representing citizens' opinion at least but strictly speaking, the associa- potential external resource for the aging neighborhood in the impacted tions are voluntary gatherings by NVCA leaders rather than official public area. In sum, Fig. 1 shows the two perspectives and variables to identify administration. Therefore, the District Association is intersecting voluntary and public roles between citizens and local government officials. After the Hanshin-Awaji Earthquake in 1995, Voluntary Disaster Man- agement Organizations (hereafter referred to as VDMO) was formally legis- lated in the law, and each municipality was required to promote VDMO to enhance mutual help in the community [36]. VDMO was expected to con- duct evacuation drills and raise awareness through DRR education in peace time, and initial fire extinguisher, guiding of evacuation, rescue of residents, coordination, and communication with related stakeholders in emergency cases [36]. The territory of VDMO and NVCA is overlapping, as 94.15 of VDMO are organized by the officers who administer each NVCA (Fire and Disaster Management Agency, 2011), therefore, NVCA is the main subject in this study. Fig. 1. Framework and the variables of the research.

2 K. Otsuyama, R. Shaw Progress in Disaster Science 9 (2021) 100141 interrelation with DRR activities in this paper as a sociocentric network case study in Japan.

3. Research ﬁeld

3.1. Demography and historical background in the city

Kesennuma is located in the Miyagi Prefecture, in northeast Japan (Fig. 2). Historically, Sanriku-coast had been famous as rich fishing grounds along with the Black Stream (Kuroshio) as remains and shell-mounds in the Stone Age had been found in Kesennuma as the archaeological evidence of human livelihood [25]. Kesennuma has a limited amount of flat land surrounded by mountains and faces the Pacific Ocean with a ria coast where large-scale tsunamis repeatedly hit the coast in 869 CE, 1611, 1793, 1835, 1861, 1896, 1933, and 1960 [40]. Fig. 3. Population transition in Kesennuma city. (Source: Kesennuma the 9th Reconstruction Report, 2014). Kesennuma City has a population of 68,232 (as of 2014). Fig. 3 shows the population transition in Kesennuma-city. The population is decreasing by 26% since 1980, while the over 65-years population is increasing more than double compared to 1980's [26]. In addition, over 65-years of population exceeded the child population in 1995, as a result, it is reported 3.2. Tsunami impacts in Kesennuma City that the entire population will be reduced by almost half of the present population in 2040 [26]. Depopulation and higher aging-population had been The city had 1356 deaths, 215 missing persons, 15,815 damaged challenges in this city since the pre-disaster. houses, and 9500 households affected by the GEJET [24]. Although most Kesennuma City has gradually become larger through mergers of small of the coastal areas were affected by the tsunami, death rate is totally differ- towns and villages in 20 and 21st centuries and formed in 2007. ent in each area (Table 2). The highest death rate is Hashikami, 4.3%, since fi fl Kesennuma currently consists of nine districts, which were former villages Hashikami is surrounded Paci c Ocean and holds a large at seacoast. Al- fl [24]. Fig. 4 shows the history of merger in Kesennuma city, and eight yel- though one community located in a at area had designated an evacuation low villages/towns and one brown district indicate present nine districts hill, tsunamis hit the evacuation place, and 93 people were killed at the in Kesennuma city. The last merger was pursued in 2009 with Motoyoshi place [12]. town, following the merger with Karakuwa town in 2006. Former If the focus of damage and loss changes at the district level, the differ- Kesennuma city, Motoyoshi town, and Karakuwa town also merged several ence is clearly shown. The highest rate is Suginoshita in Hashikami, times with neighboring villages. Each of the nine districts holds NVCA. For 29.8% due to the damage at the evacuation place mentioned above [12]. instance, a small district with a population of 2896 is divided into 13 Central parts in Shishiori and Central Kesennuma are also high, since NVCAs, while the largest district with a population of 17,387 has 67 these places have reclaimed from the salt pan after World War II [45]. For NVCAs. Kesennuma City holds nine districts whose boundaries are equiva- instance, Hama-Machi in Shishiori, low-lying area, had only 124 citizens lent to former towns/villages prior to the mergers. Table 1 shows the de- so they lost almost 10% of population, and about 20% of victims in both mography of each district in the city. communities were over 65-years old [50].

Fig. 2. Location of Kesennuma. (Source: National Map from Google Earth, and the local Map by Authors)

3 K. Otsuyama, R. Shaw Progress in Disaster Science 9 (2021) 100141

Fig. 4. History of merger in Kesennuma [24].

asked whether NVCA organized recreation, and how often they organized. Table 1 If the NVCA did not have any such activities, the point was zero and vice Demography of districts in Kesennuma city. versa. Also, they were asked whether NVCA organized evacuation training District Number Population Area Population Primary industrial in the neighborhood. Again, if the NVCA did not have any such activities, of NVCA size density work force (%) the point was zero or if the local municipality organized the event, the 2 (km ) point was 1, but if NVCA organized the events by themselves, the point Central 66 17,387 9.55 1820 2.8 would be higher. The questionnaire survey was answered by NVCA leaders Kesennuma in the city from 16th September to 13th December 2015 in each commu- Shishiori 23 5768 43.02 134 5.4 nity. Table 3 shows the number and percentage of responded districts and Matsuiwa 22 8498 34.56 246 6.5 Oshima 13 2896 8.96 323 22.2 NVCA. Total responded NVCA was 139 (62.1%) out of 221. Based on the Omose 11 6265 –– 5.8 result, follow-up interviews were also administered at Hashikami and Hashikami 13 4332 10.72 404 15.9 Shishiori districts later to identify the difference between two districts. Karakuwa 12 6954 42.32 164 16.1 The non-structured interviews District Association leaders were admin- Motoyoshi 40 10,529 106.70 99 12.8 istered based on the results of questionnaires in December 2015 and Febru- Niitsuki 21 5836 75.79 77 12.1 ary 2016. Hashikami and Shishiori were the selected districts due to the (Source: Miyagi-prefecture, 1953; Kesennuma-city statistics, 2013) asymmetric results which indicates later. The interviewer asked not only activities at NVCA levels but activities and the way of consensus building at District Associations level in the post-disaster phase. Table 2 Damage and loss data in each district in Kesennuma city. District Number of Death rate Destroyed Destroyed house 5. Results and analysis death (%) houses (%)

Central 360 1.8 7746 50.5 5.1. Correlations among variables Kesennuma Shishiori 204 2.6 3841 61.5 First, correlation among variables were examined in collected 125 sam- Matsuiwa 119 1.4 2287 36.1 ples (Table 4). Excluded lower correlation coefficients, p-values with more Oshima 34 1.0 1418 37.9 fi Omose 57 0.9 1269 35.4 than 0.400 of correlation coef cients were indicated in the table (ten vari- Hashikami 208 4.3 2718 61.5 ables out of twelve). Neither NVCA activities such as Local Festival, Tradi- Karakuwa 101 1.3 2133 28.2 tional Performing, Tea Party,andGroup Health Exercise were strongly Motoyoshi 137 1.2 3264 28.5 correlated to DRR activities, nor Networks with Social Welfare, other Niitsuki 14 0.3 465 9.1 NVCAs, Universities,andNPO/NGO. The data shows that DRR Education in (Source: Kesennuma Statistics, and Committee of Hashikami History, 2013). NVCAs were relatively related with Local Hazard Mapping, Evacuation Drills, and Local DRR Planning. Although correlation coefficients do not tell us the cause and effect, the results imply that general results have no strong corre- 4. Methodology and data collection lation with external stakeholders' interventions. Looking at the detail of DRR Education, the overall average in 125 sam- Our methodology consisted of two phases: questionnaire survey for ples was 2.13 points, and the Hashikami District is the highest average (3.3 NVCAs in Kesennuma City, and non-structured interview to District Associ- points) whereas Shishiori is the lowest (1.54 points). The two districts were ation leaders in two selected districts. The questionnaire which developed the first and second highest death and destroyed houses rates in the city, based upon the variables has five-grade evaluation. The 12 variables were but why were the points of DRR education different? These two districts translated into the questions that asked the frequency of activities and the were chosen as a case study to identify why there were such contrasting re- method of organizing events in NVCA. For example, the respondents were sults. To begin the further analysis, an understanding of the areas' basic

4 K. Otsuyama, R. Shaw Progress in Disaster Science 9 (2021) 100141

Table 3 Details of data collection. District Number of NVCA Collected samples Valid response Respondent rate (%) Date (in 2015)

Central Kesennuma 66 32 30 45.5 September, 14 Shishiori 23 11 11 47.8 December 11–13 Matsuiwa 22 15 14 63.6 September, 17 Oshima 13 8 7 53.8 September, 16 Omose 11 7 5 45.5 September, 12–16 Hashikami 13 11 10 76.9 December 13–16 Karakuwa 12 12 11 91.7 September, 11 Motoyoshi 40 27 23 57.5 September, 19–25 Niitsuki 21 16 14 66.7 October, 10 Total 221 139 125 56.6 –

Table 4 Correlation among variables. Variables

Tea parity 1.000

Group Health Exercise 0.487⁎⁎⁎ 1.000 DRR Education 0.266 0.102 1.000 Local Hazard Mapping 0.320 0.271 0.416⁎⁎⁎ 1.000 Evacuation Drills 0.220 0.133 0.439⁎⁎⁎ 0.503⁎⁎⁎ 1.000 Local DRR Planning 0.287 0.222 0.456⁎⁎⁎ 0.466⁎⁎⁎ 0.565⁎⁎⁎ 1.000 Network with Social Welfare 0.372 0.244 0.285 0.301 0.220 0.253 1.000 Network with other NVCAs 0.324 0.202 0.187 0.244 0.316 0.290 0.468⁎⁎⁎ 1.000 Network with Universities 0.090 0.084 0.200 0.146 −0.020 0.248 0.175 0.264 1.000 Network with NPO/NGO 0.201 0.080 0.224 0.204 0.241 0.148 0.273 0.324 0.417⁎⁎⁎ 1.000 ⁎⁎⁎ p <0.01. demography, the impact of the tsunami, and historical characteristics of 1896, called the Meiji tsunami, killed 433 residents [12]. Relocation to both districts are reviewed. higher places had been promoted in the impacted areas, as a result, some communities were able to avoid the impact of the 1933 tsunami 5.2. Two major contrasting communities: Hashikami and Shishiori district [48]. One of the countermeasures at the time was the relocation of fishing communities to a higher place within Hashikami District territory, which Hashikami is a coastal district with a population of 4332; the major pri- enhanced communication between inland fishing communities and groups mary industries are fishing and agriculture. Hashikami had the highest [41]. Despite the lessons from previous events and the many evacuation death rate (4.3%) and ratio of homes destroyed (61.5%) by the tsunami be- drills before 2011, the tsunami killed 208 people in the district. cause of its location, a low-lying plain along the coast (Fig. 5). Throughout Shishiori is in the northern part of Kesennuma, with a population of its history, Hashikami has suffered from tsunamis—in fact, the tsunami in 5768. The district consists of three areas: mountain, ria coastal, and industry development zones for manufacturing. Shishiori had the second-highest mortality rate of 2.6% (204 people), and the destroyed home ratio was 61.5% (the same as Hashikami). Fig. 6 shows the industry development zone in Shishiori a month later from the tsunamis in 2011. Damage and loss were concentrated in the densely populated industry development zone since the area is a low-lying, which had been reclaimed and transformed from salt flats into an industrial area after World War II [23]. Con- sequently, tsunami waves climbed inland by more than 2 km.

5.3. Contrast results of correlation coefﬁcients

Table 5 shows the correlation coefﬁcients in Hashikami District and Table 6 for Shishiori. Relatively strong correlations were observed in DRR Education with Local Hazard Mapping and Network with NPO/NGOs.Activi- ties in NVCA in Hashikami such as Local Festival, Tea Party,andGroup Health Exercise also correlated each other, while networks with other stakeholders seemed limited. On the other hand, the results in Shishiori indicated that Network with Other NVCA, Universities,andNPO/NGOs were highly related with Local DRR Planning.Interestingly,Traditional Performing in Shishiori strongly correlated with DRR Education and external networks such as Other NVCA, Universities,andNPO/NGOs. One of famous Tradi- tional Performing in Shishiori is Tra-mai (Tiger Dance) since the pre- Fig. 5. Topography of Hashikami and Shishiori districts. (Source: Authors based on disaster phase, and the role and meaning of Tra-mai discussed in the next the Map of National Land Information Div., National Spatial Planning and Regional section. In sum, a difference between two districts is external interventions Policy Bureau, MLIT of Japan). to the district.

5 K. Otsuyama, R. Shaw Progress in Disaster Science 9 (2021) 100141

association, an architecture group, a local retail association, and so on. The council conducted more integrated decision making in the district, and, consequently, they compiled a “Hashikami Recovery Proposal” on their own and submitted it to the city government [14]. The core principle of this proposal is shifting from dependence on government to self-initiative and, accordingly, the proposal asks for recommendations, rather than for the fulﬁllment of requests. The proposal focuses on the segregation of the roles: what Hashikami District can do by themselves, and what needs support by the city government. For instance, district roles suggest that each NVCA in Hashikami should prepare its disaster risk reduction plans, update hazard maps, and manage an evacuation center on their own [14]. There- fore, the planning process functions not only for consensus building at the district but enhancing DRR education through the consensus. Shishiori also had a District Association in the district, although decision making on recovery had not expanded to the whole district at the time in 2015. The reason is a lack of collective actions as a whole district due to the unequal impacts by tsunamis which concentrated on the low- lying industrial zone. Densely populated factory zones attracted newcomers less than a century ago; consequently, the multi-stakeholder involvement was not formed in the pre-disaster phase. The interviewee is one of new- Fig. 6. Industry development zone in Shishiori on April 6, 2011. (Source: Google comers from the other part of Kesennuma because his elderly brother had Earth). succeeded their parents' home and land, therefore he needed to move out from the original place. 5.4. Hashikami district recovery proposal by local-intra organizations Summarizing the results, Hashikami and Shishiori had similar impacts from the tsunamis, although the local activities in the post-disaster phase The root cause of the two different pathways in post-disasters in have been asymmetrical. As mentioned, Takeuchi [41] noted that the relo- Hashikami and Shishiori were extracted through the non-structured inter- cation caused by the Meiji tsunami had increased the mixture of residents in view with the leaders of District Associations. The Hashikami District Asso- Hashikami, inﬂuencing the neighborhood interaction as a district. This ciation was expanded into the Hashikami District Planning Council, which means that residents' interactions and communications have been observed included 46 local citizen-based organizations such as a women's since the Meiji tsunami in Hashikami, which formed the present-day

Table 5 Correlation coefﬁcients in Hashikami district. Variables

Traditional performing 1.000

Local Festival 0.766⁎⁎ 1.000 Tea Parity 0.547 0.672⁎⁎ 1.000 Group Health Exercise 0.668⁎⁎ 0.439 0.138 1.000 DRR Education 0.380 0.118 0.493 0.116 1.000 Local Hazard Mapping 0.461 0.244 0.563 0.263 0.760⁎⁎ 1.000 Evacuation Drills 0.108 0.483 0.338 −0.104 0.197 0.059 1.000 Local DRR Planning 0.570 0.580 0.102 0.500 0.352 0.142 0.367 1.000 Network with Social Welfare 0.208 −0.179 0.209 0.280 0.527 0.467 −0.123 0.069 1.000 Network with other NVCAs −0.121 −0.447 −0.662⁎⁎ 0.271 −0.193 −0.015 −0.540 0.093 0.373 1.000 Network with Universities 0.555 0.084 −0.062 0.745⁎⁎ 0.364 0.541 −0.327 0.386 0.375 0.509 1.000 Network with NPO/NGO 0.554 0.162 0.327 0.477 0.825⁎⁎ 0.587 0.213 0.498 0.721⁎⁎ 0.047 0.587 1.000 ⁎⁎ p <0.05.

Table 6 Correlation coefﬁcients in Shishiori district. Variables

Traditional performing 1.000

Local Festival 0.242 1.000 Tea Parity 0.558 −0.376 1.000 Group Health Exercise 0.136 −0.329 0.532 1.000 DRR Education 0.633⁎⁎ −0.098 0.425 0.060 1.000 Local Hazard Mapping 0.188 0.380 0.113 −0.271 0.083 1.000 Evacuation Drills 0.404 0.396 0.042 −0.226 0.604⁎⁎ 0.604⁎⁎ 1.000 Local DRR Planning 0.897⁎⁎ 0.189 0.625 0.140 0.726⁎⁎ 0.371 0.605 1.000 Network with Social Welfare 0.175 0.067 0.242 0.185 0.256 0.725⁎⁎ 0.370 0.353 1.000 Network with other NVCAs 0.690⁎⁎ 0.274 0.330 0.041 0.678⁎⁎ 0.430 0.565 0.806⁎⁎ 0.641⁎⁎ 1.000 Network with Universities 0.650⁎⁎ 0.400 0.306 0.057 0.370 0.516 0.430 0.748⁎⁎ 0.641⁎⁎ 0.923⁎⁎ 1.000 Network with NPO/NGO 0.746⁎⁎ 0.526 0.196 0.000 0.289 0.433 0.361 0.671⁎⁎ 0.493 0.822⁎⁎ 0.917⁎⁎ 1.000 ⁎⁎ p <0.05.

6 K. Otsuyama, R. Shaw Progress in Disaster Science 9 (2021) 100141

Hashikami. Such a historical background contributed to organizing the dis- for insights on recovery, Kesennuma Municipal Government for kind assis- trict council with intra-connected local resource groups and compiled the tance, and SEEDS Asia for the support. District Recovery Proposal.

6. Discussion References

6.1. Local-multi-stakeholders' involvement in Hashikami [1] Aizawa T, Saito R, Tsuchitori T, Umeya K. Research on intangible folk cultural assets in Kesennuma (I) [Kesennuma shi niokeru Mukei Minzoku Bunkazai no chosakiroku (I)]. Local-Multi-stakeholders' involvement contributes to connecting the Study Educ Region Design. 2013;4:23–41 (in Japanese). dots as a whole community in Hashikami District. Network with external [2] Ando M. Hanshin Awaji Earthquake reconstruction plan and the project. (Hansin-Awaji- Daishinsai Fukkou-Toshi-Keikaku-Jigyou-Machidukuri). Tokyo: Gakugei Syuppan; 2004 (in resources might be an entry point that Ichiko [49]pointed out that planners Japanese). and research need ‘accompany-with approach’ to encourage and engage [3] Arnstein S. A ladder of citizen participation. J Am Plann Assoc. 1969;35(4):216–24. local citizens for community-driven recovery planning. He also described [4] Bekkers R. Participation in voluntary associations: Relations with resources, personality, and political values. Polit Psychol. 2005;26(3):439–54. the entry point that one NPO started consultation of health for disaster vic- [5] Berke P, Cooper J, Aminto M, Grabich S, Horney J. Adaptive planning for disaster recov- tims in temporary shelters in Hashikami District. Eight months later from ery and resiliency: An evaluation of 87 local recovery plans in eight states. J Am Plann the disaster, the main concern was shifted to sheltering to recovery, then Assoc. 2014;80(4):310–23. [6] Breton M. Neighborhood resiliency. J Community Pract. 2001;9(1):21–36. the NPO asked support from the expert [51]. A key takeaway from this [7] Browne KE. Standing in the need: Culture, comfort, and coming home after Katrina. study is that Hashikami District demonstrated self-driven approach with Austin: University of Texas Press; 2015. local-multi-stakeholders as a driver, and external networks play a supple- [8] Burby RJ. Making plans that matter: Citizen involvement and government action. J Am – fi Plann Assoc. 2003;69(1):33 49. mental role as a supporter. This nding supports an argument of Davidson [9] Cabinet Office, Government of Japan. Disaster Management in Japan (Nihon no Saigai [13] that citizens' participation in the decision-making process is far more Taisaku). Cabinet Office, Government of Japan; 2015 (English and Japanese). important than participating in planned implementation. [10] Chandrasekhar D, Zhang Y, Xiao Y. Nontraditional participation in disaster recovery planning: Cases from China, India, and the United States. J Am Plann Assoc. 2014;80 (4):373–84. 6.2. Newly reclaimed land and geographical history in Shishiori [11] Comerio M. Disaster recovery and community renewal: Housing approaches. J Policy Dev Res. 2014;16:51–68. Another key takeaway is the geographical history underlying the disas- [12] Council of Hashikami History. Memory of the Great East Earthquake and Tsunami [Fukuyou-no-Ki]. Miyagi: Council of Hashikami History; 2013 (in Japanese). ter risk. Consecutive works of Traditional Performing in Shishiori [1,43,44] [13] Davidson CH, Johnson C, Lizarralde G, Dikmen N, Sliwinski A. Truths and myths about showed that the Tra-mai and an event of portable shrine, a traditional reli- community participation in post-disaster housing projects. Habitat Int. 2006. https:// gious ritual, have continued more than 600 years in the district, though the doi.org/10.1016/j.habitatint.2006.08.003. [14] District Planning Council in Hashikami. Proposal for Recovery in Hashikami district motivation of the event were lower for the newcomers who lived in the (Hashikami-Chiku machidukuri-keikaku teigensyo). Miyagi: District Planning Council in newly reclaimed land [44,46]. Umeya [46] clearly indicated that the pa- Hashikami; 2014 (in Japanese). rade of portable shrines omitted the reclaimed land because the area had [15] Folke C. Resilience: The emergence of a perspective for social-ecological systems analy- ses. Glob Environ Chang. 2006;16:253–67. been salt pan, as a result, he concluded that the motivation for newcomers [16] Ganapati E, Ganapati S. Enabling participatory planning after disasters. J Am Plann and over 18 to 19th generations who lived inner land in Shishiori were fun- Assoc. 2009;75(1):41–59. damentally different. The case study obviously extracted that historical per- [17] Harashina Y. Citizen participation and consensus building. [Shimi- sanka to Goi-Keisei]. Tokyo: Gakugei syuppansya; 2005 (in Japanese). spective would be a critical factor to understand local natural hazard risks. [18] Harris JL, Doerfel ML. Interorganizational resilience: Networked collaborations in communities after superstorm Sandy. In: Jones EC, Faas AJ, editors. Social Network Analysis 6.3. Limitation and the way forward of Disaster Response, Recovery, and Adaptation. Oxford, United Kingdom and Cam- bridge, MA, USA: Elsevier; 2017. p. 75–91. [19] Iiduka T. Governance in disaster recovery- affected local government challenges A limitation of this research is that the article focuses only on the con- [Shinsai Fukkou ni okeru hisaichi no governance-hisai jichitai no fukkou kadai to text of Japan and its snapshot of the recovery process as a case study. In torikumi]. Tokyo: Ashi-shobou; 2013 (in Japapnese). this study, based on Japan's background, we focused primarily on percep- [20] Inagaki F. The roles of reginal coordinators in Chuetsu Earthquake. J Assoc Rural Plann. 2013;32(3) (in Japanese). tion of the NVCA leaders four years later after the event. The leaders are [21] Iuchi K. Planning resettlement after disasters. J Am Plann Assoc. 2014;80:413–25. usually replaced every two or three years, and the perception might be [22] Jones EC, Faas AJ. An introduction to social network analysis in disaster contexts. In: changed or transformed even in the same NVCA. Regardless of the limita- Jones EC, Faas AJ, editors. Social Network Analysis of Disaster Response, Recovery, and Adaptation. Oxford, United Kingdom and Cambridge, MA. USA: Elsevier; 2017. tion, this research indicated asymmetric pathways at grassroots level as p. 3–9. an exploratory case study to look at potential variables which promote [23] Kawashima S. History of a port –Ahistoryoffisheries in Naiwan in Kesennuma, Miyagi. – DRR activities in the post-disaster phases. For the way forward, panel (Wanshi-niokeru Kisuiiki Miyagi-ken Kesennuma-Wannnai no Suisanshi kara). Nenpou-Himoji-Siryo Kenkyu. 2014;10:115–25 (in Japanese). data should be examined to identify the long-term recovery process as lon- [24] Kesennuma City. Impacts of the Disaster (Higaino-Jyokyo). http://www.kesennuma. gitudinal research. The DRR activities, NVCA activities, and external net- miyagi.jp/sec/s009/020/020/020/1300452011135.html; 2017. (in Japanese). works might be expanded or declined after ten years since the event. Accessed 30 October 2018. [25] Kesennuma-city Board of Education. Our Kesennuma-city Version 1999 [Watashitati no Such a follow-up research is needed for better understanding of the long- kesennuma shi. Kaiteiban]. Miyagi: Kesennuma-city Board of Education; 1999 (in term recovery and adaptation to the local risks. Japanese). [26] Kesennuma-city Disaster Reconstruction Department. The 9th Disaster Reconstruction Report. Kesennuma-city Disaster Reconstruction Department. (in Japanese). Retrieved October 9, 2015 from http://www.city.kesennuma.lg.jp/www/contents/1387349750 Declaration of Competing Interest 814/; 2014. [27] Kodama Y, Ogiso S. A study on the actual situation and problem of the activity of sup- fi port between the residents in recovery public housing by the great east Japan earth- The authors declare that they have no known competing nancial inter- – fl quake. Urban Housing Sci. 2017;99:128 33 (in Japanese). ests or personal relationships that could have appeared to in uence the [28] Kweit MG, Kweit RW. Citizen participation and citizen evaluation in disaster recovery. work reported in this paper. Am Rev Public Adm. 2004;34(4):354–73. [29] Lee J, Franser T. How do natural hazards affect participation in voluntary association? The social impacts of disasters in Japanese society. Int J Disast Risk Reduct. 2019;34: Acknowledgments 108–15. [30] Ministry of Internal Affairs and Communications. What is Jichikai and Chonai-kai? This study was funded by the Program for Leading Graduate Schools [Jichikai, Chonai-kai towa]. Tokyo: Ministry of Internal Affairs and Communications; 2013 (in Japanese). Global Survivability Studies Program, Kyoto University. The authors [31] Nakagawa Y, Shaw R. Social capital: A missing link to disaster recovery. Int J Mass greatly appreciate Mr. Hiroshi Usui and other local leaders in Kesennuma Emerg Disasters. 2004;22(1):5–34.

7 K. Otsuyama, R. Shaw Progress in Disaster Science 9 (2021) 100141

[32] Oikawa Y. Institutional response in education sector in Kesennuma city. In: Shaw R, ed- Community by the activating Jishu-bo.]. Tokyo: Social Welfare Benefit Promotion Soci- itor. Disaster Recovery: Used or Misused Development Opportunity. Tokyo: Springer; ety; 2011. 2014. p. 89–116. [43] Tsuchitori T, Aizawa T, Umeya K. Research on intangible folk cultural assets in [33] Okada N. Adaptive process for SMART community governance under persistent disrup- Kesennuma (II) [Kesennuma shi niokeru Mukei Minzoku Bunkazai no chosakiroku tive risks. Int J Disast Risk Sci. 2018;9:454–63. (II)]. Study Educ Region Design. 2015;6:52–62 (in Japanese). [34] Reconstruction Agency, Japan. The status of recovery from the Great East Japan earth- [44] Tsuchitori T, Aizawa T, Umeya K, Shoji Y. Research on intangible folk cultural assets in quake and tsunamis and Actions [Higashinihon Daishinsai karano fukko jokyo to Kesennuma (III) [Kesennuma shi niokeru Mukei Minzoku Bunkazai no chosakiroku torikumi]. Tokyo: Reconstruction Agency, Japan; 2020 (in Japanese). (III)]. Study Educ Region Design. 2016;7:75–83 (in Japanese). [35] Shaw R, Ishiwatari M, Arnold M. Community bousai. [Community-based Disaster Risk [45] Umeya K. The Research for Intangible Folk Culture Property in Shishiori-Chiku. Miya- Reduction.]. In: Federica R, Ishiwatari M, editors. Learning from Mega disasters: Lessons Shimbun Database; 2012 Retrieved December 6, 2015 from http://mukeidb.cneas. learnt from the Great East Japan Earthquake [Daikibo-saigai kara manabu higashi- tohoku.ac.jp/staticresource/PDF2012s-all/mukei-note-2012-s01.pdf (in Japanese). nihon-daishinsai karano kyokun]. Washington, DC: World Bank; 2014. p. 61–2(in [46] Umeya K. At the margin on the filed note: Impresson on the intangible folk cultural as- Japanese). sets in Kesennuma since the great disaster [Field note no yohakuni: Daishinsai ikou no [36] Shoji C. Chonaikai and Jishu-bosai-soshiki. In: Yoshiwara, editor. Substratum of Kesennuma Mukei Minzoku Bunkazai no Chosa Zakkan]. Kindai. 2020;121:1–37 (in Community-based Disaster Risk Reduction [Bousai Community no Kisou]. Tokyo: Japanese). Ochanomizu Syobou; 2011. p. 87–108 (in Japanese). [47] Wilson AG. Community Resilience and Environmental Transitions. London: Routledge; [39] Stevenson JR, Conradson D. Organizational support networks and relational resilience 2012. after the 2010/11 earthquakes in Canterbury, New Zealand. In: Jones EC, Faas AJ, edi- [48] Yamaguchi Y. Tsunami and Village (Tsunami to mura). Tokyo: Miyai Shoten; 2011 (in tors. Social Network Analysis of Disaster Response, Recovery, and Adaptation. Oxford, Japanese). United Kingdom and Cambridge, MA. USA: Elsevier; 2017. p. 161–75. [49] Ichiko Taro. Thinking Disaster Recovery from Community-based Risk-communication [40] Sugawara D, Minoura K. Inundation areas and deposits of the 2011 of the pacificcoast Activities － From the Pre-disaster planning for recovery in Tokyo － (in Japanese). of Tohoku and the 869 Jogan earthquake tsunami. (2011-nen Tohoku-chihou Journal of Japan Society for Disaster Recovery and Revitalization. 2020;15(Special Taiheiyouoki-Jishin tsunami to 869-nen Jyokan-Jishin tsunami no Shinsuiiki to Issue):151–8. Taisekibutsu). J Geol Soc Jpn. 2013;119:1–17 (in Japanese). [50] Tani Kenji. Distribution of the number of deaths and the death rate on the Great East [41] Takeuchi T. Fishery Village and New Life-Basic Research on Kesennuma bay. (Gyoson to Japan Earthquake (in Japanese). Research Report on Department of Geography, shinseikatsu-Kesennuma-wan Chiku kiso cyousa). Miyagi: Kesennuma board of Education; Saitama University. 2012;32:1–26. 1959 (in Japanese). [51] Ichiko Taro. Planning with local communities: Recovery and housing reconstruction [42] Tamasato E, Shimoda H, Otsuki K. Jishubousaisoshiki kasseika niyoru fukushi commu- supporting process in Suginosihta, Kesennuma (in Japanese). Disaster Recovery and Re- nity saisei no kadai to tenbou. [Revitalization and Foresight of Social Welfare vitalization Review. 2017;8(1):28–35.