Risk assessment and evaluation

RISK ASSESSMENT RISK EVALUATION

The overall task of risk management must include both an estimation of the magnitude of a particular risk and an evaluation of how important to us the risk is. The process of risk management therefore has two parts:⁴

a) Risk Assessment. The scientific quantification of risk from data and understanding of the processes involved.

b) Risk Evaluation. The social and political judgement of the importance of various risks by the individuals and communities that face them. This involves trading off perceived risks against potential benefits and also includes balancing scientific judgements against other factors and beliefs.

In order to understand a risk and to compare different risks, scientists and economists usually try to quantify it. This is done by gathering data on the effects of various hazards that cause the risk and on the basis of statistical analysis, predicting the probability of future events. The identification of causes, effects and the understanding of the processes of disaster occurrence are critical to the assessment of future risks.

The accuracy of risk quantification depends to a considerable extent on the amount of data available. The number of events on which information is available has to be large enough to be statistically significant. In addition the quality or reliability of the data has to be adequate. These factors all pose problems for the risk assessor who has to identify 'confidence limits' or range of doubt over any future risk estimations offered. Some risks are easier to quantify than others. The risks of the effects of minor floods and small earthquakes are easier to predict than catastrophic ones because they have happened more often and there is more data on their occurrence. Likewise the recurrence of droughts may be predicted on the basis of historical experience. On the other hand, risks of events that have not yet happened, such as the melt-down of a nuclear reactor for instance, have no past statistics and so have to be estimated from probabilities and forecasts.

Data collection

Collecting data on disasters is not straightforward and the systematic study of disasters is a relatively young science, so the quality of data available for risk estimation is considerably lower than that available for assessing other types of risks, like medical risks or engineering failures.

In disasters, like war, information is an early casualty. There are many large disasters that have happened this century in which, after all the confusion, it is still not known with any certainty how many people were killed let alone accurate estimations of financial losses, physical damage or disruption to the economy.

The detailed investigation of individual disasters that occur is now seen to make a major contribution to disaster mitigation efforts in a large number of countries. Flood prevention planners in Brazil, for example can learn a lot from a detailed analysis of a major flood in Bangladesh. The analysis of flood statistics world wide can help define risk levels and characteristics of return periods of floods in individual locations where specific data is scanty. The United Nations has been at the forefront of investigating and reporting disasters to the international community, through its various agencies like DHA, FAO, UNCHS, UNESCO and others.

Detailed surveys of disaster effects can identify risk factors and establish relationships between hazard and vulnerability. For example the systematic survey of earthquake damage can establish that one type of building was more badly damaged than other types - i.e. the vulnerability of one building type is greater than another. People who live in the more vulnerable building type are more at risk from a future earthquake than the others. The importance of studying the effects of hazards in order to understand risk and to make effective decisions on risk mitigation should be clearly understood. To understand risk it is necessary not just to study the casualties, but also to study those people who were not affected. Risk needs to be defined in terms of the probability of the effects and the proportion of the total population affected.