Real Risk

have historically been subject to adverse weather conditions and,

in a statistical sense, are more likely to face similar problems in

the future. Vulnerability analysis mapping, on the other hand,

commonly used since the late 1980s, combines the probability of

exposure to adverse conditions in a certain area with some

measure of ‘coping capacity’ by people living in that area. Simply

put, coping capacity refers to a population’s ability to manage a

crisis, either in the short or longer term. It depends on such

factors as level of savings and liquid assets, diversification of

production and other economic activities, mobility, and access

to alternative employment, usually through seasonal migration.

For a given area, exposure to risk for a point in time can be

expressed as the current relative level of risk compared to its

historical average – for a drought-prone area, one might use the

standardized precipitation index (SPI) for example.

Measures of

coping capacity are, on the other hand, usually expressed in qual-

itative terms, by type of livelihood system. A ‘typical’ household

in a traditional transhumant pastoral system, for example, is

assumed to have certain attributes with respect to income diver-

sification, access to alternative earning opportunities, etc.

Obviously, once one has established, for a given area, the relative

level of vulnerability compared to others (and, therefore, mapped

this area as having ‘high’, ‘medium’ or ‘low’ vulnerability), one still

has to account for changes over time in the variables determining

exposure to risk and extent of coping capacity. National food secu-

rity monitoring and early warning systems in many countries do a

good job of mapping long-term, or structural, vulnerability. When

historical patterns repeat themselves, they provide a good guide of

where vulnerability to food insecurity is being heightened by a disas-

ter. However, adverse weather can be atypical, and the variables

underlying coping capacity, such as the terms of trade between

cereals and small ruminants, for instance, or food prices and

seasonal labour opportunities in neighbouring countries, can change

in unexpected ways or faster than anticipated.

Unless monitoring systems can capture and integrate this kind

of information in a timely fashion at the national level, there can

be significant differences between ‘vulnerability maps’ and the

actual geographical distribution of the consequences of food inse-

curity, as measured by malnutrition rates, for example. During

the recent Niger crisis of 2004-2005, long-term vulnerability maps

provided a good guide to some areas where food insecurity was

indeed acute, but they missed significant southern areas with

higher rainfall. In such areas, very high population density and

drastic changes in land tenure reduced the households’ capacity

[

] 27

2001-2004

1996-2000

1991-1995

1986-1990

Africa

Asia-Pacific

Europe

Latin America

Near East

Average number of food emergencies

to produce their own food; faced with the record high prices of

2004-2005, they became even more food insecure than popula-

tions living in more marginal parts of the country, but where

extensive cultivation can still provide a partial hedge against risk.

Overall, population in a given area is considered vulnerable if

long-term exposure to risk (such as drought) is higher than long-

term coping capacity. In the short-term, however, risk exposure

and coping capacity can significantly depart from historical

average levels. Timely monitoring and analysis of the determi-

nants of risk and coping capacity, by main ecological or livelihood

system, allow for a ‘redrawing’ of relative vulnerability maps upon

which one can base localized surveys or even emergency relief

interventions. This type of dynamic risk mapping requires a good

understanding of the determinants of risk and of coping capac-

ity by livelihood system, combined with rapid information

gathering and analysis capability. This remains a challenge for

many countries, but technological advances in weather monitor-

ing, satellite data acquisition and processing, and data collection,

communication and analysis are easing the task, efficiently and

at relatively low cost.

With support from the European Commission, FAO has devel-

oped an analytical tool for vulnerability analysis and early warning

– the GIEWS Workstation allows the analyst to combine on one

electronic desktop country maps and geographical features, satel-

lite imagery including vegetation index and rainfall estimates, and

geographically referenced data such as land use and crop distri-

bution, population and price statistics, etc.

The GIEWS workstation is now being introduced for use at the

national level. In Ethiopia, FAO, in collaboration with the World

Food Programme, is providing the workstation to the Central

Statistical Authority and to the Ministry of Agriculture. The tool is

to be used in connection with ground surveys designed to calibrate

and validate medium or high resolution satellite imagery to derive

much more precise estimates of area cultivated and agricultural

production, timing of planting and yields. It will also integrate basic

food and livestock prices, as well as district-level information on

incomes and other household parameters. Since the workstation is

designed to allow for synchronization of data between separate

units (ten will be installed in the country), information exchange

will be faster and more accurate. With the help of such a tool, coun-

tries like Ethiopia will be able to specify not only which areas are

traditionally more vulnerable, but also, and in real time, how the

weather and economic conditions of each growing season modify

the risk landscape of the main ecological zones of the country.

1992-2004

1986-1991

100

Human-induced (mainly conflict)

Natural (mainly drought)

Mixed

% of crises for which primary cause was

Source: FAO

Food emergencies by region

Source: FAO

Primary causes of food crises