CHAPTER 1.

INTRODUCTION

o Had Zimbabwe had a credible, accurate ENSO forecast in early 1991, and seriously acted upon it, it could have saved tens of millions of dollars during the 1991/92 drought.

o With a credible, accurate forecast in early 1991, Southern Africa may not have experienced the costly emergency conditions that allegedly threatened millions with severe food shortages and famine, and sidetracked its efforts toward economic development.

The report begins with a closer look at ENSO in Southern Africa and the historical context and influence of drought in the region. A detailed account of "what was" the actual situation during the 1991/92 drought is then provided, followed by an assessment of "what might have been" the situation, identifying the potential benefits of ENSO information use for food security in Southern Africa.

The Southern African Development Community (SADC)¹was the original focus of the study. In the early 1990s, the Republic of South Africa (RSA) was not then part of SADC. With the end of the apartheid regime, the new South Africa was admitted to SADC as a member in 1994. Thus, RSA was not a focus of this study but is noted as it affected the rest of Southern Africa.

1.1. Problem Statement

o Farmers in Zimbabwe (normally a maize exporter) had diversified their crop portfolios away from maize in part because of lower producer prices being offered by the government. As a result, regional food stocks fell to precariously low levels and pockets of hunger and famine already existed in 1991.

o It is unclear whether the 1992 drought should have been referred to as the worst drought in meteorological or agricultural terms, or in terms of its potential societal and economic impacts on the inhabitants of Southern Africa.

o As early warning signs of food-related problems were slowly appearing in 1991, several Southern African countries were involved in economic structural adjustment programs fostered by international donors and financial institutions. These programs were designed to reduce stagnant growth, budget deficits and inflation, and had direct effects on both regional and national food security.

In 1992, Southern Africa -- a region encompassed by SADC and the Republic of South Africa (RSA) (Figure 1.1) -- experienced what has been called the worst drought in 80 years by some observers and the worst in a century by others. An estimated 30 million of the region's 100 million inhabitants were directly affected (Battersby, 1992; Chiledi, 1992; Harsch, 1992). Cereal production in the SADC region fell from a normal annual output of over 11 million metric tons (referred to as tonnes hereafter), to less than half of this amount, and cereal imports increased from an annual average of 1.6 million tonnes to over 6 million tonnes (SADC/FSTAU, 1993). Countries in the region were adversely affected to varying degrees. See Table 1.1 for general statistics on four of the SADC countries affected: Botswana, Mozambique, Zambia and Zimbabwe.

	Other Background Information
Botswana	food deficit; drought prone; land-locked; first country removed from "Lesser Developed Country" group
Mozambique	food deficit; avg. per capita food consumption far below nutritional requirements; Marxist state; in civil war from early 1980s - 1992; RENAMO rebels supported by RSA
Zambia	50% of pop. in urban areas; land-locked; only 2.5 million ha of 42 million arable hectares are farmed; several SAPs attempted since 1970s; presidential election in 1991
Zimbabwe	land-locked; normally an important regional exporter of maize; strong government-sponsored social programs through 1980s

Table 1.1 Background statistics: Botswana, Mozambique, Zambia and Zimbabwe (From Ghatak, 1995; IMF, 1996)

Drought in Southern Africa has been linked to the El Niño-Southern Oscillation (ENSO) (Preston-Whyte and Tyson, 1988; Rasmusson, 1991; Tyson, 1986). El Niño refers to the recurrent, quasi-periodic appearance of warm sea surface water in the central and eastern equatorial Pacific. The Southern Oscillation is a see-saw pressure pattern between Darwin (Australia) and Tahiti. When pressure is high in Darwin, it is low in Tahiti and vice versa. During ENSO events, Southern Africa tends to experience drier than normal conditions.

The SADC region was plagued by this drought situation on the heels of poor rainfall and crop production in the 1990/91 growing season. In addition, commercial farmers in Zimbabwe (normally a maize exporter) had diversified their crop portfolios away from maize in part because of lower producer prices being offered by the government. As a result, food stocks fell to precariously low levels and pockets of hunger and famine already existed in 1991 (Sachikonye, 1992). Thus, it is unclear whether 1992 should have been referred to as the worst drought in meteorological or agricultural terms, or in terms of its potential for societal and economic impacts on the region's inhabitants.²

In early 1992, it became clear to the global community that agricultural and meteorological drought had sharply reduced regional food production. Various national, regional, and international humanitarian organizations responded over the next several months to improve the deficient regional food supply situation. One major response was the importation of over 11 million metric tonnes of food. The amount, timing, and source of these imports are shown in Figure 1.2. Obtaining food for distribution in the region was made difficult because the region's traditional surplus food producers (and, therefore, regional food exporters) -- South Africa and Zimbabwe -- suffered major drought-related crop reductions (UN FAO, 1996). Thus, food-deficit countries in the region faced a condition of double jeopardy. Food production within their own borders was unusually inadequate and there was no food readily available to import from their neighbors.

As early warning signs of these food-related problems were slowly appearing in 1991, several countries in the region were involved in economic structural adjustment programs (ESAPs) fostered by the International Monetary Fund (IMF), the World Bank and other international donors. These programs were designed to reduce stagnant growth, budget deficits and inflation, and had direct effects on both regional and national food security. For example, as a part of structural adjustment, several governments in the region ended their subsidies for food, thereby causing food prices to increase sharply. This caused a rise in the number of people dependent on food aid in some countries, even before it was clear there would be extensive crop loss as a result of the emerging drought in early 1992. Thus, one could easily argue that the net effect of ESAPs in the SADC region was to increase the level of vulnerability of these societies to the impacts of drought (Sachikonye, 1992).

Assessing the Drought Response

o The 1991/92 Southern African drought may be a case where decision makers viewed famine as an event rather than a process. They were looking for irrefutable evidence of a famine before taking action.

The official story now goes as follows: Once the information about an impending drought-related regional food shortage became irrefutable in early 1992, appropriate action was taken by SADC's food security unit, the international community, non-governmental organizations (NGOs), and in some cases by national food security units. According to the Congressional Research Service, "This drought demonstrated how crucial a well-coordinated response can be in averting famine. The U.S. government focused its efforts on moving quickly and coordinating a response to the drought with other donors" (Dinoto, 1993: 16).

The result was the avoidance of widespread famine throughout the region and accolades for just about everyone involved. But, could the response have been better? That is, could it have been more timely and, therefore, more efficient and less costly? Did information exist about the possibility of drought in the region, such as information about an ENSO event in the central equatorial Pacific Ocean and its possible teleconnected link to drought in the region, that might have prompted a timelier response from the SADC countries and the international community, and a different set of hedging strategies and tactics?

It is important to note that famine can be viewed as an event or a process. Those responsible for releasing grain supplies or providing financial assistance often see famine as an event. Before taking action, they require hard evidence that a famine is indeed underway (or soon to be underway). They rely on quantitative indicators such as number of hunger-related deaths or the number of people being fed by emergency food programs. In contrast, others view famine as a process. For example, individuals and organizations concerned with nutritional problems may become concerned about a possible famine situation when they observe reduced crop yields, increased grain prices, reduced rainfall, or declining nutritional status of infants -- conditions which if left unchecked are likely to lead to a famine event (Glantz, 1990). The 1991/92 Southern African drought may be a case where decision makers viewed famine as an event rather than a process.

ENSO Forecasts in the Region

o As early as March 1991, there were a number of indicators (including, but not limited to an ENSO forecast) that regional food production and therefore supplies would fall far short of the needs of the inhabitants of Southern Africa. SADC's Regional Early Warning System (REWS) issued a warning that regional food stocks were dwindling to such a low level that there would be very little exportable grain within the region.

o SADC reviews contain conflicting information on when an ENSO forecast became available in the region.

As early as March 1991 and without any consideration given to the possibility of adverse meteorological conditions during the 1991/92 growing season (October through April), SADC REWS issued a warning that regional food stocks were dwindling to such a low level that there would be very little exportable grain within the region because of the poor harvest the year before (1990/91). ENSO information (including forecasts) was but one of a growing number of indicators that regional food production and therefore supplies would fall far short of the needs of the region's inhabitants.

SADC reviews contain conflicting information on when an ENSO forecast became available in the region. For example, the SADC/Food Security Technical and Administrative Unit (FSTAU) (1993) assessment of the response to the drought noted that "National, regional and international agencies working closely together provided the first advanced warning in September 1991that El Niño conditions were indicating a possible problem" (emphasis added). Elsewhere, however, it has been noted that ENSO information became available to SADC only in December 1991(e.g. Rook, 1994; UN FAO, 1996). For their part, some ENSO researchers have claimed that credible information about ENSO (including a forecast) existed as early as March 1991, although such information had not been widely disseminated.

Inconsistencies such as these raise several questions. Was ENSO information actually available to key people in the affected region before late 1991? Was there a disconnect between the information that was available to some researchers and the ENSO-related information that was actually available for use by regional or international food security decision makers? Was there an understanding of the ENSO phenomenon in the SADC region? Or about the reliability of teleconnections between ENSO episodes in the equatorial Pacific and drought in Southern Africa? Was ENSO information available to those individuals making decisions about food security and structural adjustment activities affecting food security?

1.2. Methodology

In this study, we use the case-scenario method ⁴ to assess how ENSO information was used by various actors in response to the 1991/92 drought and how that response might have been different had ENSO information been available and accessed in March 1991. We first recreate the 1991/92 Southern African drought situation. We identify the actors involved, the physical conditions of the drought, the ENSO information available in the region, the decisions made regarding food security, and regional food stocks at the time of the drought. We assess whether and how ENSO information was used by various actors in responding to the possibility of drought-related food problems.

We then consider whether and how behavior might have been influenced had ENSO information been available to food security decision makers in March 1991. Using information obtained from reports on the drought and from open-ended questionnaires and interviews, a set of responses were identified that in theory could have been taken with an earlier forecast of an impending drought situation. We then identify existing constraints on the actual use of that information. This allows us to distinguish between the theoretical value and the actual value of ENSO information in this particular drought situation.

Foot Notes

Back to text

2. See Wilhite and Glantz (1985) for a discussion on the difference between agricultural and meteorological drought.

Back to text

3. Retrospective assessments must be carefully and objectively reviewed in order to determine the correct sequence of events. As President John F. Kennedy remarked following the Bay of Pigs debacle, "There's an old saying that victory has a hundred fathers and defeat is an orphan." The avoidance of famine in Southern Africa in 1992 could be just such a victory.

Back to text

4. The case-scenario approach has been successfully applied to an assessment of long-range forecast for the West African Sahel (Glantz, 1977), for the spring wheat region of Canada (Glantz, 1979), and for an El Niño forecast for the Peruvian anchoveta fishery (Glantz, 1981).

Back to text