UPDATE: April 2011

This note is to briefly explain why Oxfam has revised its view of its 2009 forecast of a likely rise in the number of people affected by climate-related disasters. 

Two years ago Oxfam did an analysis of the number of people affected by past climate-related events. From this analysis we made a forecast that by 2015 it was likely that the average annual number of people affected by climate-related disasters would be 375 million and this represented an increase of 54 per cent compared with the average figure of the decade 1998-2007.

Having reviewed the data and the method we used to analyse the data we are no longer confident in the specific approach we used, given the nature of the data, as described below, and so we are no longer confident of this specific forecast.

Our concerns are essentially two-fold: the limitations of the data and the limitations of the way we analysed the data.

The limitations of the data: The data we used was from EM-DAT, the international disaster database managed by CRED, the Centre for Research on the Epidemiology of Disasters. This is a respected set of data which captures, amongst other things, the number of disasters that has happened and the number of people affected.

Like most sets of data it has its limitations. In EM-DAT’s case one of the key limitations is that the more recent data is more reliable. This was well known to us and we dealt with this in our analysis by using double exponential smoothing on the data (see the explanation of how we carried out the analysis and catered for the data’s limitation).

However one limitation we did not know at the time and have only recently discovered is that the number of people affected by climate-related disasters in China is unusually low in first half of the 1980s when compared with subsequent years. It is highly likely that this is the result of under reporting from China. Given the size of China’s population its disaster figures are significant when considering the global picture. The likely impact of this on our forecast may mean that we started our analysis with data from 1980 that was likely to be unreasonably low and therefore made our projected increase in the number of people affected artificially high.

We now know that there has been a significant increase in reporting from many other countries over the same period due to better information and communication.

Limitations of our analysis: Though we were clear in our explanation of the way we came to our conclusion that “different forecasting models could lead to different results,” what we should have done, particularly given the high volatility of the data especially in 2002, was run these different models to help determine the degree of confidence in the conclusion we came to.

It does not mean though that there will not be an increase in the numbers of people affected by disasters in the future. There is evidence to point to this likelihood.

According to EM-DAT the number of climate-related disasters has increased by 35 per cent from the 1990s to the 2000s. Floods have increased by 50 per cent over the same period. Munich Re’s database of global disasters also shows an increasing trend in climate-related events. It is not possible to know how much of this is an increase in events or better reporting of events and neither can we assume that an increase in events will lead to an equal increase in numbers affected.

Population growth means that there is a likelihood of more people being affected by climate-related events.

The number of people exposed to some climate-related events is also on the increase. According to the forthcoming United Nations Global Assessment Report 2011 the number of people exposed to floods and cyclones has doubled between 1970 and 2010. It is likely that many people who are ‘exposed’ to hazards are not affected due to measures that protect them, such as flood defences for example. However in many vulnerable countries investment and efforts to reduce significantly the risk of disasters is sorely lacking leaving millions exposed and likely to be affected by disasters.

Finally as the effects of climate change become more apparent there are more frequent and more intense climate-related events according to the Intergovernmental Panel on Climate Change Fourth Assessment Report. 

Oxfam is in the process of further research into the number of people affected by climate-related events and will be publishing the results of this research in due course.