Development and disaster management practitioners use risk analysis or assessment methods when drawing up project plans and making operational decisions. Risk assessment analyses the risks that threaten a project and the options for reducing those risks. Hazards and vulnerability assessments (discussed in Sections 3.5 and 3.6) form a major part of an overall risk analysis, though they can be carried out separately. It is perhaps most helpful to see risk analysis in a broad sense, as an interpretation of all kinds of data on hazards, vulnerabilities and capacities. In practice, the difference between risk analysis, hazards analysis and vulnerability analysis is often blurred. The terms ‘risk’ and ‘vulnerability’ are used quite loosely, with some overlap. Many ‘vulnerability’ assessments include wider risk analysis, whilst many ‘risk’ assessments focus on vulnerability.
Project risk analyses cover various kinds of risk, including factors within the project’s control (e.g. poor design or performance management); external factors in the wider policy and institutional environment that are outside the project’s control (e.g. institutional weaknesses, lack of political support); and other risks that will have a damaging impact on project beneficiaries (e.g. environmental hazards, conflict, sudden changes in commodity prices).
Risk analysis may be based on quantitative or qualitative data, or a combination. Quantitative risk analysis requires extensive and accurate ‘hard’ data, and tends to focus only on those aspects of risk that can be most easily measured. It often needs to be combined with other forms of information and analysis to give a comprehensive view. Advanced scientific knowledge and computer modelling enable sophisticated quantitative risk analysis, but also require a high level of resources and technical skills and may not in any case be required; in the fire risk assessment in Vientiane (Case Study 3.1), much of the data needed was collected by visual surveys and a relatively straightforward scoring system was used. Qualitative analysis uses descriptive scales for the likelihood and magnitude of risks. It is mostly used for initial screening, where the level of risk does not justify fuller assessment or there is insufficient data for more quantitative analysis. It often takes the form of a probability/impact matrix (see Figure 3.2).
A fire risk assessment in the city of Vientiane in Laos identified seven key risk factors and gave a numerical value to each to arrive at a total risk score for each geographical unit surveyed.
|Risk factor||Total score|
|Building material type||25|
|Sources of flammable material||15|
|Fire-fighting scenario (availability of water and manoeuvring space for fire-fighters)||15|
|Quality of electrical wiring||5|
There was a sub-set of quantifiable features within each of the seven categories. Again, each carried a numerical score. For example, under ‘fire history’, there were four categories of risk: high (4 incidents of fire recorded during the past 5 years – score 10), moderate (3 incidents – score 5), low (2 incidents – score 3) and very low (1 incident – score 1).
P. Sounnalath et al., ‘Fire Risk Assessment in Vientiane Lao PDR’ in Proceedings: Regional Workshop on Best Practices in Disaster Mitigation: Lessons Learned from the Asian Urban Disaster Mitigation Program and Other Initiatives, 24–26 September 2002, Bali, Indonesia (Bangkok: Asian Disaster Preparedness Center, 2002), http://pdf.usaid.gov/pdf_docs/pnadk776.pdf, pp. 97–102.
OECD, Disaster Risk Assessment and Risk Financing: A G20/OECD Methodological Framework (Paris: OECD, 2012), http://www.oecd.org/gov/risk/G20disasterriskmanagement.pdf, p. 43.
One common limitation of risk analysis is that it does not take a broad view of human vulnerabilities and capacities, tending to focus on more visible and quantifiable elements, such as buildings and physical or financial assets and human lives. It is possible to capture less visible aspects through more qualitative, participatory risk analysis, and the results of such exercises can be valuable in understanding local perceptions and priorities (see Case Study 3.2: Risk mapping among East African pastoralists).
A team of researchers developed a simple but systematic approach to classifying and ordering the sources of risk faced by pastoralists in arid and semi-arid districts of southern Ethiopia and northern Kenya. The aim was to find a robust participatory method that was less costly and time-consuming than full surveys. There were two stages in the method: risk identification and risk ranking. The first stage was achieved using an open-ended questionnaire. The researchers emphasised to the pastoralists that they could list as many problems as they wished, and should identify these through discussions amongst themselves. The second stage used a simple numerical ranking method to group risks in order of severity. Risks thought to be equally severe could be ranked equally. After they had done the ranking, informants were asked to discuss each risk in turn, explaining how they used to deal with the problem, or why they no longer could, and how they would like to overcome it. Assessment of the incidence of risk was based on the proportion of participants who identified it. Severity of risk was assessed using a mathematical calculation that translated the informants’ perceptions into a simple risk scale. Findings could be plotted on maps to identify areas and groups at risk. Disaggregation by age, gender, wealth and other socio-economic characteristics was also possible. The method was tested in the field over six months, involving 120 groups (59 groups of women, 61 of men). The responses identified 15 major sources of risk, ranging from availability of food and water to banditry. The most frequently mentioned problems were insecure access to food and water, livestock disease and access to health clinics. Food and water shortage were the only risks mentioned by a majority of informants, indicating that the extent of the other risks varied considerably across the region and its population, even though some (e.g. malaria, conflict) were certainly severe in places.
K. Smith, C. Barrett and P. Box, ‘Participatory Risk Mapping for Targeting Research and Assistance: With an Example from East African Pastoralists’, World Development, 28 (11), 2000.