Social networks and challenges in government disaster policies: a case study from Bangladesh

Rabiul Islam*, Greg Walkerden

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

15 Citations (Scopus)


This study analyses recent disaster management and climate change policies in Bangladesh, evaluating the ways in which they do, and do not, take households' social networks into account. The analysis of the policy documents is qualitative: drawing out explicit themes and implicit acknowledgements through close reading of the text. The evaluation of their treatment of households' social networks is based on a combination of field research, including disaster practitioner and policymaker interviews, meetings with local journalists and local government officials, and NGO workshops, and a review of the literature. The findings show that despite the importance of social networks in disaster resilience and recovery, these networks are not given adequate emphasis in government policies. Organisations' linking social networks (i.e. governments' links with foreign states, donors, and others) are emphasised in policies, but households’ bonding networks (households' relationships with immediate family members and relatives), bridging networks (households' relationships with neighbours and friends), and their local linking relationships (notably with NGOs and local government) are largely ignored. This study argues that existing disaster management policies should be revised, with substantial local community input, to better leverage local social capital.

Original languageEnglish
Pages (from-to)325-334
Number of pages10
JournalInternational Journal of Disaster Risk Reduction
Publication statusPublished - Jun 2017


  • Bangladesh
  • Cyclone
  • Disaster policy
  • Resilience
  • Social capital
  • Social networks

Fingerprint Dive into the research topics of 'Social networks and challenges in government disaster policies: a case study from Bangladesh'. Together they form a unique fingerprint.

Cite this