FOOD EMERGENCIES, FOOD SECURITY AND ECONOMIC PROGRESS IN DEVELOPING COUNTRIES

Mwita Rukandema and A.A. Gürkan^[44]

The paper argues that frequent and often protracted disasters which lead to food emergencies in developing countries are a major threat to their economic and social progress and long-term food security. In Section 1 key concepts such as a disaster and a food emergency are defined, followed in Sections 2 and 3 by a review of the current food emergency situation and trends and causes over the past 20 years. This review indicates that Sub-Saharan Africa is the worst affected sub-region. In Section 4 potential impacts of various types of disasters on developing economies are discussed, along with some examples. Discriminant analysis is used to explore the inter-relationships between food security and a range of socio-economic variables. This analysis reveals that the more food insecure group of developing countries is characterised, inter alia, by low levels of socio-economic development and economic growth rates, high incidence of food emergencies and high food import bills relative to export earnings. Finally, levels of agricultural productivity across the developing regions are assessed and sub-Saharan Africa's debt burden is highlighted. Section 5 concludes and draws policy implications of the paper's various findings. Major policy implications are that developing countries, with the assistance of the international community, should implement measures to reduce vulnerability to disasters and to mitigate their impacts and enhance their economic performance. For the poorest and heavily indebted countries, creditors need to drastically reduce or write-off the debts in order to make more resources available for investment.

1. Introduction

FAO's Global Information and Early Warning System (GIEWS) classifies a country as facing a food emergency when a disaster-induced shortfall in its aggregate food supply relative to its consumption requirements in a given year cannot be fully covered by the country's own resources and, therefore, it needs external food assistance. In the absence of such assistance, the population or a significant proportion of it would be seriously undernourished or, in extreme cases, face famine.

The shortfall may be caused by a natural or man-made disaster, or a combination of both. A disaster is defined as a serious disruption of the functioning of a community or a society causing widespread human, material, economic, or environmental losses which exceed the ability of the affected population to cope on its own (UNISDR, 2002). Natural disasters are caused by events (or natural hazards) such as earthquakes, volcanic activity, drought, floods, landslides, tropical cyclones, pest and disease infestations, etc. On the other hand, man-made disasters are caused mainly by wars and financial/economic crises. However, what may appear to be a natural disaster may actually be the indirect result of human activities which harm the environment. For example, river flooding may be due to heavy soil sedimentation arising from serious soil erosion in the river's catchment area, which in turn is due to deforestation and land degradation. Human activities that harm the environment result in increased vulnerability of local communities to disasters. Furthermore, sometimes a natural disaster may strike a country already suffering from a man-made disaster. For example, over the past few years, Afghanistan has suffered simultaneously from serious droughts and civil war.

It is now widely acknowledged that natural disasters, particularly droughts and floods, are increasing both in frequency and severity. There is also evidence to show that man-made disasters, particularly armed conflicts, have increased steadily over the past two decades. Developing countries, mostly the poorest, bear the brunt of the world's disasters and suffer disproportionately from them because they lack the means both to prepare for them and to repair or rebuild shattered infrastructure afterwards. Governments must divert scarce resources from much needed development programmes to reconstruction activities and to meet escalating food import bills, thereby compromising economic and social development and long-term food security. Coupled with the high costs of servicing external debts, the added cost of disasters is crippling for poor countries.

2. Current food emergency situation

Table 1 indicates that as of July 2003, some 36 countries around the world were affected by food emergencies, including 23 in Africa (64 percent), 7 in Asia (19.4 percent), 4 in Latin America (11.1 percent) and 2 in Europe (5.5 percent). In Africa, the dominant cause is civil war which affects 14 of the 23 countries or 61 percent, followed by drought (11 countries or 48 percent). In Asia, the main cause is drought, especially in Central Asia where the impact of the 2000 severe drought is still being felt and food assistance continues to be required. In Central America, the collapse of international coffee prices has left thousands of families without an income, as workers on coffee plantations have been laid off en masse. In Europe, many internally displaced families and refugees in Chechnya (Russian Federation) and in Serbia and Montenegro need food assistance to survive, following recent or ongoing civil wars. It should be noted that although a country may have recently emerged from a disaster (e.g. Afghanistan, Sierra Leone, Angola), a food emergency may persist for several years. Furthermore, several countries are affected by two or more disasters simultaneously (e.g. Somalia, Sudan, Zimbabwe, Afghanistan, Iraq), which makes the emergency quite complex.

Table 1. Countries facing food emergencies as of July 2003 (total: 36 countries)

FAO's latest (July 2003) estimates indicate that the number of people affected by food emergencies worldwide in need of food assistance stands at around 70 million, compared to an average of 64 million over the past three years.

3. Recent trends in food emergencies

Food emergencies all over the world have increased steadily since the mid 1980s (Figure 1). The figure, based on the FAO/GIEWS database, also shows that for some years (e.g. 1993, 1995, 1998 and thereafter), certain countries suffered more than one disaster in a single year.

Figure 1. Trends in food emergencies globally

Figure 2 shows that man-made disasters have been on the increase since around 1990.

Figure 2. Trends in causes of food emergencies

On the other hand, from a peak in 1984, natural disasters fluctuated considerably until around 1997 when they began to escalate. The main causes have been extreme hydro-meteorological events (mainly droughts, floods and cyclones/hurricanes).

Figure 3 shows trends in food emergencies by region. Sub-Saharan Africa has experienced the largest number over the past two decades, ranging from nine to 28. The number of people affected is currently estimated at 40 million, compared to an average of 27 million over the previous three years. The main causes have been civil wars and droughts, followed by floods. An upward trend is evident since 1996, notwithstanding a small decline in 1999. In Asia, although the absolute numbers of emergencies are relatively small, the number of people affected is quite large, currently estimated at 32 million, more or less the same as the average of the past three years. The graph shows an upward but weak trend. The main causes have been floods and droughts, with the financial crisis of 1997/98, exceptionally harsh winters in Mongolia and over a decade of civil war in Afghanistan all playing a significant role. In Latin America, the pattern is less clear, but until 2002 emergencies were caused almost entirely by natural disasters, mainly hurricanes, floods and droughts. The peak in 1999 reflects "Hurricane Mitch", which devastated several countries in Central America in October 1998. In Europe, the trend mainly reflects civil wars among Commonwealth of Independent States (CIS) countries following the break-up of the former Soviet Union, as well as the widespread drought in 1999/2000.

Figure 3. Trends in food emergencies by region

4. Impact of disasters on long-term food security and economic progress of developing countries

4.1 Natural disasters

There are two main types of natural disasters, those which develop relatively slowly such as droughts, and those with rapid onset such as floods, earthquakes, cyclones/hurricanes and landslides. The first type lends itself to sufficient early warning and preparedness, thus reducing the potential consequences, especially loss of human lives. On the other hand, sufficient early warning may be almost impossible for the second type and usually there are casualties in its wake. However, with recent advances in information technology including the use of satellites and computer-aided modelling, there is now more scope for early warning, even if still limited. For example, forecasting the timing and location of tropical cyclones has doubled warning time to 48 hours since 1990 (UNISDR 2002). From the perspective of food security and economic progress in developing countries, the following two paragraphs focus on disasters caused by three hydro-meteorological hazards, namely floods, cyclones/hurricanes and droughts, hazards that are becoming increasingly frequent and severe, with escalating economic, social and material costs.

Although the immediate physical impact of floods and cyclones/hurricanes may be confined to a relatively small part of a country such as a region or province, a large number of people may be killed, hundreds of thousands or even millions made homeless, and extensive damage or destruction inflicted on economic and social infrastructure and productive assets (roads, bridges, railways, schools, hospitals, public buildings, power lines, livestock, crops, factories, homes, etc). UNISDR (2002) concludes that in addition to the projected estimate of 100 000 lives lost annually due to natural disasters, the global cost is anticipated to exceed US$300 billion annually by the year 2050, if the likely impact of climate change is not effectively countered. The lack of capacity to limit the impact of disasters remains a major weakness of developing countries, especially the poorest which also face higher levels of disaster risk because of adverse interactions between poverty-induced human activities and the environment.

The impact of drought tends to be more widespread, covering an entire country or even a whole region as exemplified by southern Africa recently and Central Asia in 2000. However, apart from losses in crop and livestock production and sometimes sharp reductions in hydro-electricity supply, the impact on infrastructure and other physical assets is minimal. Also, loss of human life is not great thanks to the slow onset of drought-induced disasters which allows sufficient time for early warning, usually followed by reasonably adequate and timely international response. Nevertheless, governments in developing countries are compelled to spend more of their scarce foreign exchange resources to finance increased food import bills. Furthermore, since agriculture is the backbone of manufacturing industries and the main employer in developing countries, economy-wide impacts of droughts seriously slow down or even reverse economic growth. The wide fluctuation of rainfed crop production in sub-Saharan African countries is one of the major causes of instability in overall economic growth rates. Thus the increasing frequency of droughts is a major threat to economic and social development and long-term food security in developing countries. At the household level, a succession of droughts deepens poverty and increases food insecurity and malnutrition, which in turn lead to human activities that damage natural ecosystems, thereby increasing the risk of further disasters.

4.2 Man-made disasters

Two major causes of man-made disasters in the world today are economic crises and civil wars. Recent examples of economic crises include the financial collapse in Indonesia in 1997/98 and in Argentina in 2002, and the current economic crisis in Central America following a ruinous crash of international coffee prices. In Indonesia, the sudden collapse of the rupiah sent shock waves throughout the economy, resulting in a sharp rise in unemployment to over 20 million and sending close to 100 million people below the poverty line within one year (FAO 1998). This episode coincided with an El Niño-induced drought that sharply cut paddy production and raised rice import requirements in 1998/99 by nearly 50 percent, at a time when foreign exchange was extremely scarce. In Central America, thousands of families are currently depending on international food assistance following abandonment of coffee plantations, a major source of livelihood in the sub-region.

Civil wars in many developing countries today, particularly in sub-Saharan Africa, are another major threat to long-term food security and economic progress for these countries. Section 2 indicates that food emergencies are concentrated in sub-Saharan Africa and that for the majority of the affected countries (61 percent) the dominant cause is civil wars. The human and economic cost of civil wars is always enormous. In the Democratic Republic of Congo, for example, an estimated 3 million people have died since the outbreak of civil war in 1998. The dead have been mostly civilians killed by the warring factions and by diseases affecting the internally displaced who are forced to live in subhuman conditions. Others have died of starvation. Angola, Afghanistan, Sierra Leone and Sudan have been in similar situations. In general, high military expenditures on civil wars and the incalculable direct and indirect costs to economies can reverse a country's economic progress by decades.

In order to explore in more detail the relationships between food emergencies and long-term food security and economic and social progress in developing countries, discriminant analysis is employed to determine whether there are characteristics that are consistently different in those developing countries that are considered to be relatively more food insecure compared to those that are not. FAO's estimates of the proportion of undernourished in total population are used to divide the developing countries into two categories: those where the proportion is above 15 percent being considered as more food insecure than those below (FAO 2003a, forthcoming). The cut-off of 15 percent as a criterion for determining the membership into one or the other group is quite arbitrary. If "reasonable" consistencies in the differences in the characteristics of the two groups of countries are revealed by the statistical analysis, then the criterion would have proven its operational usefulness.

Studies undertaken elsewhere (World Bank 2002), suggest that conflict-related food emergencies are associated closely with economic performance, as well as with social and political factors, all of which in turn tend to be intimately linked to food security at both the national and household levels. In this study, a selection of such indicators for 92 developing countries for which estimates are available is used to explore the nature of these relationships. These are:

• the Human Development Index (2001), a composite variable representing many dimensions of the development process (see UNDP 2003);

• GDP per capita (US$ - 1995) and growth rates of GDP per capita (during the 1990s), representing economic performance;

• various measures of food imports (averages for the 1990s), representing the ability of countries to access international food markets; and,

• total number of food emergencies experienced in each country (during the 1990s).

The estimates of the proportion of undernourished are the most recent available, using data for the period centred at 2000. Apart from the Human Development Index, which also uses the most recent data, all other variables are calculated using data pertaining to the decade of the 1990s. This has been done so that it may be possible to interpret some of the associations discovered in the analysis to imply causality in an empirical sense, since the indicators represent certain processes that occur chronologically prior to the state variables which are observed at the end of the period.

Discriminant analysis is a multivariate statistical technique that creates successive independent linear combinations of the variables in such a way that the means of the linear combinations (or discriminant functions) for each of the groups are as different as possible, after controlling for the effects of the preceding functions. With only two groups, as in our case, only a single discriminant function can be extracted; with, in this case, the coefficients of the function being estimated so that the mean values of the function in each of the two groups are farthest from each other. If, moreover, it is discovered that the variances of the function in each of the two groups are much smaller than the total variance of the function, then it becomes possible to assign a high level of (statistical) confidence to claims that the characteristics of the countries in the two groups, as measured by the underlying variables, are sufficiently different from one another.

These statistical tests indicate, with a level of confidence greater than 99.999 percent^[45], that the differences between these two groups of countries are large. One way of understanding the nature of the differences is to look at the patterns of the structure coefficients or the correlations between the variables and the discriminant function itself, which Chart 1 visually displays. The variables have been ordered to reflect the nature of the relationship between them and the discriminant function. Positive structure or correlation coefficients indicate that larger values of the variables tend to be associated with larger discriminant scores, and vice versa and negative correlations indicate that smaller values of the variables are associated with larger discriminant scores, and vice versa. The signs of the coefficients, therefore, help group similar variables together, while the magnitudes determine the strength of the relationships between them. The largest positive coefficient is that of the Human Development Index (with a value of 0.96) and the largest negative coefficient belongs to the total number of food emergencies declared throughout the 1990s. Through the structure of the relationship of each variable with the discriminant function, it is also possible to deduce the nature of the variables to one another along this discriminating dimension. This means, for example, that those countries that have a lower Human Development Index tend also to be ones that experience greater food emergencies, and vice versa.

Chart 1. Structure of the discriminant function in terms of the underlying variables used to characterize the countries

A fuller and more meaningful discussion of the findings could be facilitated if we conducted the analysis using the vector of means of the seven variables as they apply to the two groups of countries categorized on the basis of the proportion of their population that is deemed to be undernourished. Chart 2 displays the group means of the variables in the same order as in Chart 1. It is obvious that those variables that have positive correlations with the discriminant function also exhibit larger means for the group of countries that are relatively more food secure (i.e. with lower a proportion of their population undernourished), and those that have negative correlations exhibit smaller means for the same group of countries. The substantive nature of the consistency in the differences is obvious:

• Countries in the food insecure and more vulnerable group have lower levels of economic and social development, smaller initial GDPs, have developed more slowly and depend less on international trade to provide for their food needs - the opposite applies for the more food secure group of countries;

• Countries in the food insecure group face a greater number of food emergencies and have a larger share of food import bills in total merchandise imports than the more food secure ones;

• Countries in the more food secure group have a larger share of imports in total apparent food consumption as they have access to larger resources for imports as evidenced by a smaller share of food imports in total merchandise exports.

Chart 2. Averages of the seven variables for the two groups of developing countries

Thus the general thrust of these findings is that the state of household food security at the national level (as measured by the proportion of population that is undernourished) at any moment appears to be crucially dependent on level of economic development and performance, on ability to access food imports to supplement what can be produced domestically and the degree to which falling into food emergencies can be avoided. Perhaps a new and significant insight that deserves reiteration is a reminder that self-sufficiency in food is not necessarily a good indicator of food security. The evidence provided here indicates that food self-sufficiency may indeed be the result of factors such as limited import capacity that constrain the ability of developing countries to obtain some of their food through international trade.

4.3 Other food security consequences of natural and man-made disasters

Repeated or protracted emergencies also lead to cumulative effects that deepen poverty and undermine development capacities. A diversion of resources from productive investment results, inter alia, in a stagnant agricultural sector, rising unemployment and increasing national indebtedness. The issues of low agricultural productivity and indebtedness are briefly highlighted below.

Low agricultural productivity

Figure 4 shows per capita cereal production in Africa, Asia and Latin America between 1980 and 2002. In Africa, per capita production has not only been very low (125-175 kg/year), but has also stagnated at around 150 kg/year. This mainly reflects limited use of productivity-increasing agricultural technologies (improved seeds, fertilizers, irrigation, tractors, etc.) but also the impact of frequent and often protracted disasters on the continent. In Asia, per capita production is relatively higher (245-288 kg/year) and shows an upward trend until 2000. Widespread use of improved agricultural technologies and intensification of production are the main determining factors. With relatively high levels of agricultural production and export earnings, Asian countries are generally better able to cope with or avoid food emergencies. In Latin America, per capita production is also relatively high (230-280 kg/year) and has trended upwards since 1993. Like Asia, Latin America generally has a capacity to cope with food emergencies, except for the poorer countries of Central America such as Guatemala, Nicaragua, Honduras and El Salvador which have suffered a succession of natural hazards (El Niño, Hurricane Mitch, Hurricane George, earthquakes) and are now gripped by the coffee crisis.

Figure 4. Trends in per capita cereal production, 1980-2002 by region

External debt burden

Balance-of-payments difficulties are particularly acute for sub-Sahara African countries, largely due to high external debts relative to their ability to repay. Figure 5 shows per capita external debt of sub-Saharan Africa from 1980 to 2000. The debt rose from US$175 per person in 1980 to a peak of US$400 in 1995, and stood at around US$300 in 2000. The decline since 1996 probably reflects debt relief under the Highly Indebted Poor Countries (HIPC) initiative. Nevertheless, the debt burden is unsustainably high for countries whose per capita income averages less than US$450 per annum, with the poorest at around US$100 (Ethiopia, Burundi), and about 50 percent of their total population living below the poverty line (World Bank 2002, AfDB 2003). While a number of factors, including unfavourable terms of trade and high international interest rates, played a major role in raising the level of indebtedness, it is highly probable that repeated or protracted disasters contributed a significant part. Servicing these debts is a serious constraint to long-term food security and economic and social development of the sub-region.

Figure 5. Sub-Saharan Africa per capita external debt (US$)

Source: World Bank, FAO.

5. Conclusion and policy implications

This paper argues that the increasing frequency and intensity of disasters is a major threat to long-term food security and economic progress of developing countries, particularly of the poorest which lack the capacity to cope with their impact. Scarce resources are diverted from development programmes to rehabilitation and reconstruction activities, military budgets and escalating food import bills. The worst affected sub-region is sub-Saharan Africa, which currently accounts for 64 percent of the total number of countries suffering from food emergencies globally. As regards natural disasters, developing countries need to be assisted to establish or run effective early warning systems that make use of advances in information technology, plus the necessary technical training. These countries themselves need to promote water management practices such as flood control, small-scale irrigation, water harvesting, watershed management and land-use planning to reduce vulnerability to droughts and floods and to mitigate their impacts. Civil wars are particularly concentrated in sub-Saharan Africa, which currently accounts for 82 percent of the global total. Enhanced economic performance is indicated in this paper to be a crucial requirement for prevention or mitigation of conflict-induced food emergencies. Therefore, poor countries need to be effectively assisted to achieve sustained economic growth, diversify their economies and achieve higher agricultural productivity. International assistance needs to target the poorest (and conflict-prone) countries. Moreover, it is essential that developed countries open up their markets to allow expansion and diversification of exports from developing countries. Furthermore, the crushing debts of the poorest countries need to be drastically reduced or cancelled to make more resources available for domestic investment.

References

African Development Bank (AfDB) 2003. African Development Report

FAO. 1998. GIEWS Special Report: FAO/WFP Crop and Food Supply Assessment Mission to Indonesia. October 1998.

FAO. 2001. Food Emergencies and Outlook - Conference Supplement to Food Outlook.

FAO. 2003a. The State of Food Insecurity in the World 2003, Rome, forthcoming.

FAO. 2003b. The Impact of Disasters on Long-Term Food Security and Poverty Alleviation - Policy Implications. Committee on World Food Security. May 2003.

UNISDR (UN International Strategy for Disaster Reduction). 2002. Disaster reduction and sustainable development: understanding the links between vulnerability and risk to disasters related to development and environment. Background Paper for the World Summit on Sustainable Development, Johannesburg, 26 August-4 September 2002.

UNDP (United Nations Development Programme). 2003. Human Development Report 2003. Millennium Development Goals: A compact among nations to end human poverty, published for UNDP by Oxford University Press, New York.

World Bank. 2002. World Development Indicators.

World Bank. 2003. Breaking the Conflict Trap: Civil War and Development. Policy Research Paper by Paul Collier.