This chapter gives a brief account of the historic evolution and current context of food security information systems, technical aspects of the FSIEWS methodology contained in this handbook, including its specific advantages, structure, mechanisms and didactic approach. Finally it gives a definition of food security that is widely accepted today throughout the world.
The handbook is divided into three main parts. The first part defines the main aspects of the FSIEWS method (definitions, description of national and international contexts, aims of the system, etc.). The second part, by far the longest, describes the different stages in setting up a FSIEWS, the constraints encountered and the possible solutions. The third part attempts to place the system in a wider more flexible and lasting context, and to identify how its role may evolve over time in a global food security context.
This handbook is aimed at middle managers in developing countries since they are the ones who actually implement the FSIEWS and keep them running smoothly, whether as part of a statistics monitoring system in one of the food security sectors, as one of the many other actors in food security (in the public, voluntary or private sector), or as providers or recipients of information. Their commitment is key to the long-term success of a FSIEWS. An enabling national policy, supported when necessary by external technical assistance, are the two other important aspects, also closely linked to the work of the professionals to whom this book is addressed.
The text of the handbook is a distillation of a number of technical papers produced by national and international experts and technicians. It was given its final form at FAO after thorough technical revision by the appropriate services and all the participants.
The advantages of this methodology are as follows:
Although the FSIEWS methodology was not developed by specialists it is the fruit of many years' work in the field by national and international teams. This handbook is a distillation of their experience and sets out the essential stages in defining and setting up a national system (specific to each country) for monitoring food security and early warning, the characteristics of which are as follows:
In order to define and set up a monitoring system, a detailed analysis of food security in the country, including any specific problems, must be carried out.
This definition is in keeping with the three classic aspects of food security: availability of staple foods, stability of supplies and access for all to these supplies; but it also contains the idea of adapted food, what is known as the "biological utilization" of food.
Forecasting is the foundation of all warning systems. It must be applied to the four areas of food security (availability, stability, access, biological utilization), giving decision-makers enough time to react to the warning, but with a high enough degree of reliability (as a general rule the more long-term the forecasts the less reliable they are) to avoid false alarms. All forecasts have a probability rating (calculable or not), which gives a good idea of their reliability.
In the four food security areas, forecasting techniques are generally based on information gained from monitoring the situation and on diagnostics using:
Forecasts are sometimes carried out by the organizations in charge of monitoring information (those responsible for monitoring markets often analyse price trends), but are just as often carried out by other bodies (meteorological services, often under the ministry of transport, can be responsible to carry out agrometeorological forecasting), or by a specific body gathering information on food security and early warning. The relevant body should enter the results of the forecasts (indicators, warning signals, etc.) in a food security control panel that is constantly updated and to which decision-makers have constant access. The aim of the control panel, including warning signals, is to steer food security efforts and supply the various media aimed at decision-makers at all levels (bulletins, radio programmes, updates, fax, etc.).
The purpose of the control panel is similar to a dashboard in a vehicle in that it presents different kinds of information that can be analysed in order to anticipate events in each of the monitoring and forecasting areas. For example, the pilot of an airplane can assess the risk of crashing by taking in at a glance the petrol gauge close to zero, the warning signal showing the engine temperature is too high, and the spirit level showing the plane is at too steep an angle. Similarly, national authorities will be concerned about a sudden rise in grain prices in the markets, the closure of certain roads because of flooding, an increase in the death rate, and a warning signal set off by large population movements. Even though the data may be fairly imprecise, the combination of these elements is much more important than knowledge (even precise knowledge) of one or other of these elements. Forecasts and decisions are often made on the basis of an accumulation of (sometimes biased) indices. The use of indices (even imperfect ones) helps to draw attention to situations that need to be examined more closely. In a car when one of the red lights on the dashboard goes on, you stop, open the bonnet and, if necessary, carry out the repairs.
Availability of staple foods. The foods that should be monitored are the staple foods required by the poorest populations (see Part Two, Chapter I). For rainfed crops, the harvest forecasts are based on agrometeorological data, on estimates of the areas sown and by using agronomic parameters for the region and crops in question. Different methods of varying sophistication are used to calculate them. Forecasts for root or tuber crops in humid zones are calculated by estimating, often approximately, areas and yields; while animal production forecasts use other parameters (changes in the condition of grazing land, number of animals recorded at certain water holes, etc.). Import and export forecasts are reliable when importers (public or private) agree to submit their import-export schedules in countries where the customs system is strictly controlled (some island countries for example). Nevertheless, in the majority of developing countries forecasts are based on estimates regularly supplied by specialists and are fairly approximate. In countries where imports and exports have to be authorized by administrations in advance, the most reliable estimates regarding the movements of staple foodstuffs can therefore be obtained from the relevant services.
Stability of supplies. Forecasts in this area are usually based on the analysis of market trends, taking into account the socio-economic and political situation (price trends as well as changes in quantities available in markets, stocks, inter-regional transfers, etc.). In the commercial area, as was mentioned above, forecasting is often very difficult and can be fairly unreliable.
Access for all to supplies. Access is linked to the constraints of relative poverty (financial means/price of staple foods) and to physical access to these products (see below for a more detailed analysis). Monitoring indicators for poverty and retail prices are useful for medium- to long-term trend analyses but are generally not suitable for short-term forecasts and "best-guess estimates". As for nutritional forecasting (see below), sets of indirect indicators suitable for short-term forecasts of the evolution of poverty and physical access to food are generally used. Sociologists identify and monitor the strategies used by vulnerable individuals or groups in situations of short-term foreseeable crises (stockpiling of produce, selling essential equipment, seeking work away from the household, etc.). The sets of indicators that have a bearing on the short-term forecasts of poverty often overlap with nutritional ones (access and biological utilization) since sudden increased poverty almost automatically leads to a deterioration in the nutritional status of the affected group or family. The younger members of poor families tend to migrate to the towns as soon as they anticipate loss of earnings. Indeed, apparently inexplicable migratory phenomena are probably an indication of short-term problems. Moreover, a good market for used essential equipment, sold by vulnerable families, is an indication that families are in serious financial difficulty and are converting their last resources into much needed cash. Certain religious practices are also symptomatic. In order to establish a type of indicator or set of indicators, a socio-economic study of individual behaviour patterns must first be carried out in homogeneous areas and the vulnerable groups and individuals be studied (see Part 2, Chapter 1).
Biological utilization of foods. Health or anthropometric indicators are status indicators; they do not indicate changes in the nutritional status of populations and, most importantly, the most vulnerable groups. They provide information on the past nutritional levels of a population, sometimes current nutritional levels, but never future ones. Therefore indirect socio-economic indicators (or sets of indicators) are generally used to measure individuals' own perceptions of future nutritional problems. The analysis of coping strategies used by vulnerable individuals or groups in food crisis situations generally gives good indirect indicators for forecasting nutritional problems. Adults know only too well what their wasting or stunting of their children means and can foresee the consequences.
Most of the existing food-security monitoring systems are organized around the following four main pillars:
These four pillars have specific aims and set up their own means and organization. They are generally country-wide and linked to the statistics services of each of the ministries concerned.
The creation of a Food Security Information and Early Warning System (FSIEWS) by the national bodies responsible for supplying food security information should not overlook any of these aspects but take them all into account in setting up a global system. Thus the monitoring of food availability (production + imports - exports - losses) should be supported by monitoring information on both production (MAP) and foreign trade supplied by the Market Information System; the monitoring of the stability of supplies uses data mainly from the Market Information System (MIS), as well as data on the status of infrastructure and stocks; the monitoring of access to these supplies should take into account mainly social indicators (poverty, unemployment, migrations etc.); and the monitoring of biological utilization should use data acquired from health and nutritional monitoring.
The monitoring of agricultural production often focuses on cereal crops, and sometimes includes a section on animal production and/or grazing. It is usually established by the statistics services of the ministry of agriculture, which relies on regular studies in the field, in theory, carried out by the provincial administrative staff. The data is usually published annually in standard statistics publications.
Most of the crop monitoring and forecasting methods are developed around the water balance calculated during the growing season and take into account the phenological development of the plant. The agrometeorological approach has produced good results in semi-arid countries where the water deficit is the main factor limiting crop productivity. This approach gives less satisfying results2 in regions (even semi-arid ones) where:
The monitoring of animal production can use data from the veterinary services (vaccinations, abattoir figures), fiscal services (taxes), zootechnical surveys (where they exist) or monitor grazing land (mainly in nomadic areas).
The monitoring of rainfed crops is based on the following principal techniques:
Generally speaking, these tools can be used for qualitative evaluations of crop status (development, stage in the cycle, etc.), which can become quantitative depending on the availability of additional information (agronomic data, statistics on yields, historical series, etc.) and providing the information is validated. In addition, some institutions carry out essential field work to calibrate the analytical models used (see the methods used below). This work is indispensable for the proper functioning of simulation models and for testing in different conditions. NOAA satellite images can also prove useful at regional or even national level but, given the average size of farms and their patchy distribution over the territory, the calibration of NDVI5 values is a long process that takes several years to verify. The processing of Meteosat images is easier, but even in this case careful testing is necessary.
From an institutional point of view, the agricultural-production and harvest-forecast monitoring systems are usually established in two stages:
The systems are usually situated within the ministry of agriculture and the national meteorological services. In the first case, production is estimated on the basis of data recorded by district agents (phenological monitoring) and agricultural sampling surveys. In the second case, meteorological data is used and agrometeorological analyses of varying sophistication are developed. In addition, plant protection services often play a major role in monitoring the phytosanitary status of crops, and insect and locust attacks. In principle, all the services (agricultural, meteorological, dissemination, plant protection, livestock farming, water, etc.) provide, in their particular areas, information that is also included in their own analyses during the season. What is often lacking is a systematic analysis and synthesis of the information (past, present and future), procedures and methodologies used so that all the users may have access to a transparent and objective reference base.
This multi-institutional and multi-disciplinary system ensures that a progressively more sharply focused methodology is used to gather information. Using this approach it is possible to predict what areas risk poor food crops and introduce more finely-tuned surveillance techniques, that may also include sample surveys of the affected populations.
A Market Information System (MIS6) is a service, usually operated by the public sector, which involves the collection on a regular basis of information on prices and, in some cases, quantities of widely traded agricultural products from rural assembly markets, wholesale and retail markets, as appropriate, and dissemination of this information on a timely and regular basis through various media to farmers, traders, government officials, policy-makers and others, including consumers.
The system gathers, processes and disseminates information on commodity trading. The information should include monitoring data on market prices and available quantities of certain commodities (not solely agricultural and food), variations in stocks (public and private) and how goods are transported. In general, MIS also monitor imports and exports. It should be noted that the MIS play a crucial role in the dissemination of information to public and private operatives. This role has often reduced the MIS to only monitoring market prices, since the demand from traders has been very high in this area. The monitoring of available volumes, which is more difficult to implement, has often been neglected.
A Marketing Information System must carry out the following steps:
Market Information Systems are usually attached to the statistics services of government ministries (trade, agriculture etc., according to the country). In some countries, a National Institute of Statistics, usually attached to the planning ministry, is in charge of the MIS. In others, the individual ministries (for example, the ministry of agriculture for food and agricultural products, including agricultural inputs) may be directly responsible for gathering and processing data in their respective domains.
Those in charge of collecting data at food markets (quantities, prices, how the market works, origin of products, etc.) may be agents attached to different ministries (trade, agriculture or internal affairs), or to the private sector (chambers of commerce) or non-governmental organizations (NGOs). They usually work at provincial level where the data is collected, processed and passed to a central unit. Training of field staff is also (theoretically) carried out at this level. The "data-collectors" or "researchers", as they are often called, are sometimes assigned exclusively to the MIS. If they belong to a general statistics unit, they may have other surveys or administrative tasks to carry out; if they belong to the ministry of agriculture they may be involved in the dissemination of information or other activities.
In June 1999, the World Committee on Food Security (CFS) endorsed the three classes of vulnerability (or at risk of food insecurity) established by the FIVIMS7 working group. The three types of food insecurity8 are:
The poverty except when defined as the minimum needed to ensure biological survival, is generally regarded as a relative situation in monitoring systems and should only be considered as such.
An understanding of the changing vulnerability situation and what action should be taken to improve the well-being of the most deprived populations can be gained from the relevant and precise information provided by the regular collection of certain socio-economic or other indicators. This can be carried out in parallel with measurements proper of the health and nutritional status of the population. By monitoring the factors that influence situations of deprivation, a given level of poverty can be revealed. So the same income in two different families or two different countries, taking into consideration their customs, may or may not indicate poverty.
Food intake (quantity and quality), income, employment, access to resources (land, credit, etc.) and to basic services (health, education, etc.) are the most commonly used indirect indicators of poverty after a tentative evaluation in quantitative or monetary terms.
Currently, apart from certain partial or combined indicators that have been used in large data collections (income, human development index-HDI, schooling rate, health cover, etc.) there are more case studies than usable series in the databases.
The majority of developing countries do not have regular and appropriate monitoring of poverty for either marginal or vulnerable groups.
NGOs and national social services (sometimes the ministry of social affairs) may have data, collected and processed with varying regularity, regarding some areas of countries. The main purpose of this data is to monitor communities and enable populations, but it can also be very useful in monitoring vulnerable groups and in estimating local changes in food security at the level of provincial committees9.
International organizations (such as FAO) involved in developing the FIVIMS10 initiative, created in the monitoring framework of the World Food Summit11, are now endeavouring to analyse local methods of monitoring vulnerable groups in an attempt to develop a classification system that can be used to observe changes in these groups at global level.
The nutritional status of an individual and/or a population depends on all the factors that have an effect on the food/health relationship. Food and nutrition surveillance systems keep track of the nutritional status of populations and their consumption of food in order to facilitate decision making in this area. FNSS was first adopted in developing countries in 1976. The methods and aims have evolved considerably since then, and in particular since the nineties. Several countries, notably in sub-Saharan Africa, have set up FNSS to monitor and measure the impact of structural adjustment policies on vulnerable strata of the population.
Nutritionists have traditionally targeted children, but interest is increasingly being focused on the nutritional status of other age groups (adolescents, adults, the elderly) using indicators based on the body-mass index12. Budget consumption surveys give a good overview of the food and nutritional situation, but they are carried out every ten years at the most. In addition to these fairly centralized approaches, local monitoring of vulnerable groups is being developed as a routine part of the activities of governmental institutions or NGOs working at community level.
In general, the data for monitoring health and nutrition comes from five major sources13:
Food and nutrition surveillance systems are usually centralized and based in the country's capital. Monitoring is carried out by those in charge of health and nutrition (often the ministry of health) who may have decentralized offices (health centres at provincial, departmental, local or community level). The collection of basic data is the responsibility of public health officials who rely on anthropometric indicators (weight/age, height/age, height/weight ratios). Primary-school teachers are also sometimes involved in monitoring the health and nutritional state of children.
Information from other sources is sometimes used in monitoring food security:
Food security may or may not be among the overt aims pursued by those in charge of each of these systems. They often limit their aims to finding a balance between the supply and demand for some food products seen as being of prime importance. The cereal balance is an example of this approach. Others, who see economic development as a priority, focus on a "social" approach to food security, trying to correct food security problems that arise as a result of economic growth, without however integrating this dimension into their global planning. The latter therefore usually limit their food security monitoring to monitoring nutrition and vulnerable groups, preferring isolated actions targeting particular sections of the population.
Others again, particularly in areas of dryland cultivation, prefer a productive approach. They consider the monitoring of basic food crops and the corresponding harvest forecasts the principal source of information on food security in the population; for them, this information is a sufficient basis for decision making. It is also clear that chambers of commerce will be more concerned about information regarding the markets (price, quantities, stocks) than gaining knowledge of vulnerable groups, which is by definition difficult to deal with. This more profit-oriented approach is, however, quite close and complementary to the search for a balanced supply and demand.
The principal constraints encountered by current food security information systems in developing countries fall into two categories:
There are also functional constraints (found in many systems in developing countries), which must be taken into account in the organization of a FSIEWS:
The main criticisms are as follows:
The following defects may also occur:
The main functional constraints can be summarized as follows:
Rapid urbanization and the steady integration of developing countries into the world economy are the two main factors in the evolution of food security and early warning information systems.
This situation is responsible for the following changes:
The Food Security Information and Early Warning System can also be an instrument of choice in managing food crises although its main aim is to assist in the prevention of crises and avoid disastrous repercussions on food security in the short, medium and long-term.
When it works well, such a monitoring system can make a valuable contribution to managing food crises. This contribution can be broken down as follows:
The approach advocated in this handbook is designed to help national authorities to define and implement a Food Security Information and Early Warning System (FSIEWS), perfectly adapted to the constraints of each country. In practice, a FSIEWS relies on existing monitoring systems for information on the availability of staple products, the stability of supplies, access for all to these supplies and the biological utilization of food (defined as the relation between health and nutrition). In the following pages the outline of a theoretical FSIEWS is presented. The stages in setting up such a system are described in the second part of this handbook.
Experience has shown that, in order to function properly and be used correctly, a FSIEWS has to have the following attributes:
As can be seen from the following outlines, a FSIEWS should ideally be situated within the secretariat of the National Food Security Committee, be relayed by the Provincial Committees and possess two computerized monitoring instruments: a database containing information from existing databases in the area of food security, and a control panel resulting from the cross analysis of data, and forecasts of short- or long-term changes in food security, according to complex indirect indicators or estimates. (For further details on the institutional organization of a FSIEWS, see Part Two, Chapter IV, Section 4, Institutional Organization.)
The aim of such a system is two-fold: in the short-term it serves as a steering instrument for food security and is therefore an instrument of choice in preventing food crises. In the medium and long-term it should be an indispensable step in any programming and planning activity, providing planners with the data and analyses necessary for an understanding of food security.
It is, above all, an organ that enables cooperation to be established and food security activities to be proposed for entire populations, but in particular for the most vulnerable groups.
3 The NOAA satellite produces digital images from which a Normalised Difference Vegetation Index is obtained. This satellite index is largely correlated with the volume of living vegetation. In arid and semi-arid conditions the state of crops and the surrounding vegetation are closely linked.
13 National surveys are a major source of information. However, they are not strictly speaking monitoring systems given the irregularity with which they are carried out (often once every ten years). They nevertheless provide essential information about vulnerable groups at a given time and allow comparison of estimated monitoring data with statistical data obtained from these surveys, etc.