The transfer of disaster volunteer management know-how:
A case study of the Japan seacoast oil spill of 1997
Shigeo TATSUKI, Ph.D.
School of Sociology, Kwansei Gakuin University
1 Uegahara, Nishinomiya, Hyogo, 662-0519 JAPAN
Haruo HAYASHI, Ph.D.
Research Center for Disaster Reduction Systems, DPRI, Kyoto University

Based on the Kobe earthquake disaster volunteering experiences, a phase specific disaster volunteer management model was proposed at the 5thUS/Japan workshop on Urban Earthquake Hazard Reduction (Tatsuki, 1997). This working model suggested that disaster volunteer management processes consisted of three phases, emergency, development, and ending phases. The Japan seacoast oil spill of 1997 attracted high media attention, which mobilized tens and thousands of clean-up volunteers from all over the country. While local township governments were overwhelmed by the sheer number of volunteering applicants, local Junior Chamber of Commerce (JC) members were active in managing a flood of clean-up volunteers. This incident provided an opportunity to test the working model of disaster volunteer management. The similar phase transitions were clearly observed in a case of Mikuni town. However, a sudden termination of organized volunteer management activities characterized a case of Mihama town. The difference of the two towns was attributed to management know-how transfer from Kobe volunteer veterans.

Figure1. Number of Volunteers Managed at Kwansei Gakuin Relief Volunteer Center from 1/21/95 to 4/5/95
Figure 2. Phase Boundary Identification of the KGU volunteer data by Moving Standard Deviation Method

Phases of Disaster Volunteer Management

Based on his own experiences as a disaster volunteer center manager, Tatsuki(1997) proposed phases in the disaster volunteer management process, which he tentatively named emergency, development and termination phases (see Figure 1). The determination of when a phase ended and a new phase started, however, was based on intuition. Hayashi proposed a more empirical technique to identify phase boundaries. This technique is called the moving standard deviation (MSD) method. The MSD method was applied to the data in figure 1. Figure 2 illustrates the results. The MSD method identified two significantly high (more than 2 SD) spikes, one on January 31, and the other on February 7, indicating those two to be phase boundaries.

Our relief volunteer center began operation at the 100th hour after the onset of the earthquake. The first phase was characterized by a skyrocketing turnout of manpower. The wide media coverage of the disaster prompted an empathy-forming triangulation between the disaster victims and the rest of the nation. Excitement and hero-like feelings dominated the volunteer center. Every moment was a new experience in terms of disaster response management. Administratively, this was a period of organizational deployment. A paper and pencil method was first crafted in order to allocate manpower to four evacuation centers. In two weeks since the earthquake, the manpower allocation table was soon to be replaced by a computerized manpower database system which served fourteen evacuation centers surrounding the university.

According to the MSD method, a phase boundary emerged at the end of the second week, and the next phase began. The second phase was characterized by an equilibrium of high volunteer turnout. Day-to-day manpower management operation was by then routinized and newly recruited staff volunteers were fully capable of dealing with administrative tasks. This was a period of routine operation.

The MSD method identified another phase boundary on February 7th. This suggests that the routine operation phase lasted only one week before starting the final phase. Although we were aware of the decreasing turnout, we resisted the trend by devising activities that would attract and rekindle interest in our center. It was during this period that factions developed. Although both factions wanted to continue and even expand the range of services, they differed in strategy. One faction wanted to move the operation front to Kobe where there was still substantial need for material support. The other faction wanted to shift the focus to psychological care and the provision of information. Both factions were reluctant to face the reality of what is now obvious based on the present data: the termination phase had begun. Furthermore, they attributed the drop in turnout to the activities of the opposing faction. Consistently low levels of manpower turnout were interpreted as failure and resulted in a need for scapegoating. Also, by this stage, too much relief volunteer involvement seemed to prevent the shelter residents from forming governance over their own lives. This was another sign that the final phase had begun and both factions failed to recognize it.

The Japan seacoast oil spill of 1997: Comparison of Mikuni and Mihama Operations

As soon as the news of a disastrous oil spill drifting to the coast of Mikuni was broadcast on January 7, 1997, a number of Kobe veteran volunteers left for the Japan seacoast. Tatsuki (1997b) reported that the Kobe veterans transferred the following management know-how to Mikuni volunteer managers within 100 hours after the arrival of the ship wreck to the coast of Mikuni township. 1) The volunteers need to be self-reliant in terms of means of transportation, lodging, meals, and clean-up gear and should not be dependent on local government and people. 2) Local JCfs should develop and organize a volunteer management system, which should also be run by volunteers under the supervision of JC officials. 3) Volunteer managers need to be aware of the optimal balance between volunteers and local government. If the relationship is too close, volunteers may be dominated by the government and lose their autonomy to act, or interpersonal bickering may occur. If the relationship is too distant, however, no resources or information will be exchanged between the two. 4) Volunteer organizations need to form a coalition network in order to deal with local government in order to maintain their decision-making autonomy. 5) Daily broker/intermediary activities with volunteer groups and with local government are essential to maintain the network. 6) There are phases in disaster response, beginning, development, and ending phases. In order to begin the operation wisely, one needs to pre-define the conditions for operation termination long before the operation even starts.

Figure 3. Number of Oil spill Clean-up Volunteers in Mikuni and Mihama Towns

Figure 4. Phase Boundary Identification of the Mikuni data by Moving Standard Deviation Method Figure 5. Phase Boundary Identification of the Mihama data by Moving Standard Deviation Method

Figure 3 compares manpower turnout between Mikuni and Mihama operations. The MSD method was applied to the data and results are presented in figures 4 and 5. The striking difference between the two townships is when the phase boundaries occurred. In Mikuni, the first phase boundary occurred only one week after the disaster and the second phase boundary occurred at the end of the third week. In the case of Mihama, the first and only phase boundary occurred at the end of the third week. It appears that Mikuni was able to complete the first deployment phase very quickly and move promptly into the routine operation and termination phases thanks to the knowledge transferred from the Kobe veterans. On the other hand, Mihama had no such expertise and struggled much the same way that people in Kobe struggled during their own disaster. This indicates that if the disaster volunteer management process is left to its own natural course, it normally takes two to three weeks to get to the routine operation phase.

Table 1 compares the management resource characteristics of the two townships. It should be noted that Mikuni received more media attention because of the dramatic presence of the hull of the broken tanker near their shore. Therefore, most of the manpower resources, materials and money donations were attracted to Mikuni. In contrast, the Mihama JCfs received no management know-how from the Kobe veterans, fewer manpower resources, materials and money thanks to marginal media coverage. The Mihama JC pursued a ggoodh relationship with both the local government and township people. However, this relationship proved to be too close and resulted in a loss of autonomy. Because they did not recruit or train management staff from among volunteers, the JCfs remained solely in charge for the entire operation and they simply burned out.

Table 1. A phase by phase comparison of management resource characteristics.

Mikuni Town

Mihama Town


Media Event

Kobe Veterans

Volunteer Management


Routine Operation

Management OJT

Support of Other Regions

Internet Broadcasting


Rumor & Prejudice

Disaster Cowboys


Local JC

Lack of Kobe Vetfs

Relations too close with township officials and with local residents


Staff Burn-out

Lack of Management

Lack of Manpower

Immature & Dependent Volunteers

Rumor & Prejudice

This case study shows that it is possible to detect phase boundaries using the MSD method by monitoring day-to-day indices such as volunteer turnout, donations, and media coverage. The comparison of Mikuni and Mihama township responses to the same crisis indicates the importance of early know-how transfer that raises the awareness of the coming phases in advance.

Tatsuki, S. (1997a). A life-modeled social work practice with earthquake victims: Phase specific responses during crisis and post-crisis periods. A paper presented at the 5th US/Japan Workshop on Urban Earthquake Hazard Reduction.

Tatsuki, S. (Ed.) (1997b). Volunteerism and civil society. Kyoto: Kouyou Shobou.


All Rights Reserved. COPYRIGHT(C) 1998-99, Shigeo TATSUKI

Department of Social Work, School of Sociology, Kwansei Gakuin University

The Tatsuki Lab Top (English)