Humanitarian disasters triggered by natural hazards appear to be growing in severity and frequency of impact. For some, this is explained by natural cycles in atmospheric and geophysical phenomena. Increasingly, though this explanation is proving insufficient, not least as hydro-meteorological cycles are themselves influenced by the externalities of human development through anthropogenic global environmental change. Rather, a consensus has formed around the recognition that disaster events are simultaneously the products of inappropriate development policy, and vice versa. Dramatic examples of natural hazard events that have been allowed to lead on to humanitarian disasters because of inappropriate development include the Marmara earthquake in Turkey (1999), Hurricane Mitch in Central America (1998) and the Gujarat earthquake in India (2001). In these cases losses are high but spatially contained. However, there is a real chance that the economic losses caused by natural disasters might be sufficiently large in the near future that they could impact on regional, if not the global economy. From 1991 to 2000 natural catastrophes are reported to have caused losses of US$78.7 billion per annum, with disaster losses increasing fourteenfold between the 1950s and 1990s (IFRC/RC 2001).
The good news is that human security to so-called natural disasters is a rising political priority; the bad, that increased funds for disaster relief and engineering-based mitigation measures tend to mean a reduced budget for national and international development agencies. Juggling the balance between direct investment in disaster mitigation and response and longer-term development initiatives is tricky for politicians and aid agencies that tend to work to five-year budgeting timetables at most. In this environment there is a need for clear policy scrutiny based on a critical understanding of the relationship between natural disaster and development in a globalizing world. An initial response by theorists and practitioners to this challenge has been to revisit the Marxist interpretation of disaster first put forward in the 1980s (Hewitt 1983). However, the twenty-first century differs from the mid-late twentieth century in important ways: populations have grown, urban areas are burgeoning, capital is increasingly footloose, traditional cultures are increasingly modernized, the state is smaller and civil society actors are