The present study analyzes the social dynamics of disaster volunteer activities over the past 30 years, from the 1995 Great Hanshin-Awaji Earthquake, which was called the 'first year of disaster volunteers,' to the 2024 Noto Peninsula Earthquake. We have conducted collaborative practices and action research in many disasters in Japan, including the 2004 Niigata Chuetsu Earthquake and the 2011 Great East Japan Earthquake and Tsunami, over the past three decades. It was revealed that the social dynamics of disaster volunteers for 30 years could be conceptualized as the conflict between two intricately intertwined social drives, social drive for order and social drive for unstructured as the former gradually surpassed the latter. We postulated that the social, cultural, and political dynamics of these two drives could be explained by the social psychological concept of the out-group homogeneity bias of Japanese society, which would be overcome by practical interventions toward individual's social identity. Based on these findings, I also attempted to theorize regarding these social dynamics in the context of other international research findings on disaster volunteers (i.e., structure-task model). We concluded that the social drive for unstructured should be emphasized rather than order in future disasters for which the responses would expect more complicated structure of organizations and more unknown tasks than ever. These findings provided some implecations not ony for future practices of disaster volunteering but also for theoretical and methodological develpment of applied social psychological research.