The Bali negotiations on global futures raise again questions about society and risk. Some 30 years ago, in addressing the nature of risk, O’Keefe and his colleagues (1975) argued that under capitalism those most vulnerable were the most marginal in terms of livelihood. It was a crude measure of social class based on per capita GDP. There are now much more salient analyses of vulnerabilities (http://www.vulnerability.net.org). It is, however, worth returning to the argument. The argument was essentially around what we shall call ‘Type 1’ disasters. While the dominant curve of Figure 1 shows a direct correlation between the number of disasters and poverty, thus leading to discussion about marginal people and marginal places, the extremes of the graph show a different tendency. Where more income is available there is a tendency for disasters to rise quite simply because there is more property that can be damaged. At the other extreme, poor people, as measured in monetary worth, tend to be in pre-capitalist modes of production where subsistence coping mechanisms enable them to enjoy a level of resilience, thus reducing risk to disaster.