Revising Need Assessments and alternatives to food aid
Revising Need Assessments and alternatives to food aid
The presentation I’m going to make today, ‘revisiting need assesments and alternatives to food aid’, is based on the findings of two research works that I’ve conducted in the two past years in Southern Africa –Malawi, Zambia and Zimbabwe- on the response to the so called ‘Southern Africa food crisis’ of 2001-2003.
The crisis has brought a number of questions about the effectiveness and appropriatness of the response provided by the different organisations and Government bodies involved. As it is where everything starts, we will look at needs assesments, their limits and constraints, and the role they have played within the response.
Revisiting need assesments in the Southern Africa crisis
In 2002, the regional crisis was initially presented as a drought affecting production in 6 countries in the region and resulting in a food deficit of 4 millions metric tons and a food aid requirement of 1,2 million metric tons for nearly 13 millions people. It was however rapidly aknowledged that the crisis was more complex than just a drought as other factors were also often put forward, such as HIV/AIDs, bad governance and also consequences of the economical decline of the region.
The above mentionned figures are the outcome of a number of assesments conducted by NGOs, UN agencies –FAO and WFP- and Government officials from SADC countries mid 2002. Let see how they were produced…
Two main instruments
In 2002-2003, two main instruments were used to estimate the food needs in Southern Africa, one looking at national food balance and deficit, the other one at food needs at household level.
1. The estimate of food balance (CFSAM) at national level is done by FAO/WFP in collaboration with different ministries, generally agriculture and trade. It is the result of a comparison between several variables: production, predicted exports and imports, human and animal consumption. A calculation is then made to estimate the food deficit, which is generally considered to be the food aid requirement.
2. The estimation of food needs at household level was mainly done in the region using SCF’s Household Economy Approach. This methodology looks at different weatlth groups within the population (poor, middle and better off) and estimate for each group the expected food deficit (the quantity of food that households can’t produce or procure on their own).
Were these instruments relevant to Southern Africa in 2002-2003 ? Answering this question requires to pay some consideration to the context where they were used.
Context of the food crisis
For the past few decades, agriculture in Zimbabwe, Zambia and Malawi has been characterised by two main features, inherited from the colonial time and generally perpetuated since independence:
- The maize ‘mono-cropping’: maize represents the main crop for a large majority of small scale farmers and the main staple food for consumers
- The duality of the production system, in short shared between an estate-based commercial sector and a smallholding sector predominantly subsistence oriented
The production of maize by small scale farmers, principally rain fed and on small parcels, is constrained by the irregular rainfall patterns that characterise the region and the low fertility of the soil left free by the big estates. For this reason, maize production in the region has relied for long on the readily and cheap availability of seeds and fertilisers, generally supplied by the State. Besides, this uncertainty of the production has also led the States to implement some form of social protection through parastatals in charge of marketing and pricing of maize: they used to offer pan-territorial and pan-seasonal low food prices for consumption throughout nation-wide systems of subsidised sales. To be able to play this role, Governments require a policy of regulation of the food sector, including support to production, as mentioned before, but also some level of control and intervention on imports and exports. In the past two decades, economic decline and structural adjustments have led to the reduction of State intervention in the sector.
Another characteristic of the production system is that the poor soil fertility combined with the limited access to land result in poor yields and inability for many farmers to produce all the food they need to fulfil their consumption requirement. For instance, in Malawi, an average year around 40-60% of the food needs would be covered by farmers’ own food production whereas the drought increases their food expenditures from 40-60% to 70-90% of total households expenditures. People need as a consequence to find other sources of income, mainly found through work in plantations, industries and mines, either in their country or in the region.
Lastly, the agricultural sector includes also a number of commercial farmers who produce food, livestock and cash crops. They have access to mechanised agriculture and enjoy relatively high yields. Their share of national food production varies according to the strategic decision they make (type of crops to grow, area planted,…) and depends on profit opportunities.
Relevance of need assesments in this context
Estimation of the national food deficit
The CFSAM methodology and results can be questioned on a number of arguments.
- The food consumption is based on a per capita consumption estimate that is not always realistic and reliable. The estimates done for Malawi in April and August 2002, with in between a modification of the average per caput consumption represents a clear evidence of the poor reliability of these estimates (see Food Balance Sheets).
- Once the commercial imports have been estimated, the food deficit becomes automatically a food aid requirement and never suggests other sources of assistance (alternatives could be for instance cash support that will promote local production and trade)
- They often ignore the non-cereals crops although they may represent a large part of households’ food consumption (can be more than 50% in drought years)
- They ignore informal trade and cross border exchanges that are frequent in countries with poor road and market infrastructures (in 2002, estimates were 60,000MT for Zambia-Tanzania, 120000 MT for Mozambique-Malawi in 2002).
- They do not consider the decision making of traders and commercial producers who take into account the plans for relief and governmental interventions when deciding for commercial imports or crops.
- They ignore the patterns of speculation and hoarding, which have massively increased since the price liberalisation.
Even with reliable results, the CFSAM must be taken with caution for a more fundamental reason when assessing humanitarian needs: it does not assess the spatial and social repartition of the production. A good surplus situation in a given country may provide good export incomes but also hide the vulnerabilities and limited production in some outlying small-farming communities. This is an important point in dual economies like those of Southern Africa.
Estimation of the household food deficit
The food deficit at the household level depends on two variables : own food production and purchasing power (ability to cover food needs through purchase of food). The purchaing power depends on the one hand on household’s income and on food prices on the other.
The problem is that these two last variables are not really explored by the HEA approach. This is the main bias of this methodology, which focus on what people are able to buy and produce at a given time. It is obviously uneasy to predict and measure other variables such as the level of income through off-farm activities or changes in food habits. It is similarly impossible to predict the evolution of food prices along the year, including potential decreases due to Governement stabilisation interventions or depreciation due to the import of food relief.
Such a bias would not make a significant difference in countries where people are mainly or only subsistence farmers. It is not the case in Southern Africa where most subsistence farmers still have to purchase a significant proportion of their food requirement. In Malawi, an average year around 40-60% of the food needs would be covered by farmers’ own food production whereas the drought increases their food expenditures from 40-60% to 70-90% of total households expenditures.
From need assessments to responses
The estimates of food needs both at macro and micro levels can therefore be questionned because of the imperfections of the methodology used.
Now a more fundamental point must be made about the use of these assesments : when looking at what as happened in 2002-2003 for the three countries, these assesments solely provide a recommendation in terms of quantities of food to be distributed and numbers of people that should receive these distribution.
Yet a number of other tools exist to adress food insecurity and the best response is not necessarly the distribution of free food. Indeed, while it seems relevant in a situation of acute emergency (such as a sudden disaster), or in countries who have suffered war and destructuration for years, such as Liberia or Congo, it is different for the Southern Africa countries where…
- States exist and have their own capacity to handle structural vulnerabilities and transitory food insecurity
- a local and regional capacity of food production can be used against deficits.
Indeed, several mechanisms have been in place in Southern Africa to deal with food insecurity:
ß agricultural support through subsidisation of seeds and fertilisers. It can include support for winter cropping in a time of harvest failure for maize.
ß marketing food (through GMB in Zimbabwe or Admarc in Malawi) at subsidised prices, making food accessible to the poorest and to those living in marginalised areas,
ß chanelling food or cash through National Public Work Schemes (such as the one in place in Malawi) targeting the poorest and landless
ß social safety nets, such as the social welfare programme in Zambia that could have chanelled funds to reach the most vulnerable in Zambia.
Comparing 1992 and 2002
In 2002-2003, these mechanisms were not much used and most of the assistance came through free distribution of food chanelled by WFP and international NGOs. It is worth noticing that is was very different in the response to the drought in 1992:
In 1992 in Zambia, the Public Welfare System was channelling cash assistance to the most vulnerable households while NGOs were providing food relief. A national NGO, the Programme Against Malnutrition (PAM) was created as a secretariat in charge of the coordination role of the relief response and a part of the distributions.
With the exception of war-torn Mozambique, donors provided significant balance-of-payments supports to finance drought related imports and for ODA, it was actually the largest single form of support. In Zambia, donors assisted with the importation of 1 million metric tons of food, of which 10% was designated for humanitarian purposes and 90% for commercial sales .
In 2002, Zambia and Malawi got support from financial institutions to purchase cereals. In the case of Zambia, this support came relatively late in the year as a response to the GM issue and replacement of the US food that was not coming anymore at the end of 2002. In Malawi, the Government contracted a loan to purchase 250 000 MT of cereals for distribution through subsidised sales but the overestimate of food needs resulted in the inability for the Government to sell this food for more than a year in a market that was flooded with food.
In Malawi, two parallel pipelines (Government and WFP/NGOs) brought around twice the amount of food that was required with consequently not only a waste of money but also a high cost for regional and local food producers who could not sell their food. In Zambia and Zimbabwe, food aid was distributed by WFP and NGOs with not much role left to and any way no coordination with the Government activities such as the subsidised sales realised by GMB in Zimbabwe.
In 2002, a major difference with the response to the 1992 drought relates therefore to shifts in the respective role of the various actors and their relation to one another. Not only Governments and local civil society had a much greater role and ownership of the intervention but also cooperation with donors and international organisations was much closer.
A new role for humanitarian organisations ?
This shift in the role of the various players to respond to a food crisis is consistent with a broader evolution that we can notice in Southern Africa and else where.
Indeed, the removal or the reduction of State intervention in the food production is part of a broader trend related to market liberalisation and structural adjustments. It has largely participated to the decline in food production in 2001-2002. More generally, it has aggravated the fragility of small scale farming in a region where maize cropping requires a significant support in subsidised inputs, cash or credit.
On the other hand, price liberalisation has resulted in the instability of food prices, making seasonnaly food unaffordable for the poorest, especially in a bad year.
The above shows that this evolution results in increased vulnerabilities and risks of food insecurity for small farmers. Yet, at the same time, governmental safety nets are being withdrawn and replaced by external interventions by WFP and NGOs.
WFP and NGOs, especially International NGOs, are therefore replacing national mechanisms for food security and national safety nets. It brings a number of questions such as national sovereignity and accountability of the agencies but also around the objectives pursued by interventions using imported food in a potentially exporting region, such as the dumping of US food.
FAO/GIEWS: Africa Report No.1, April 2002 MALAWI 27
CEREAL SUPPLY/DEMAND BALANCE FOR THE 2001/02 MARKETING YEAR (in thousand tonnes)
Wheat Rice Coarse grains Total
Previous five years average production 2 78 2 084 2 164
Previous five years average imports 41 1 85 127
2001/02 Domestic Availability 3 45 1 901 1 949
2001 Production (rice in paddy terms) 2 61 1 770 1 833
2001 Production (rice in milled terms) 2 40 1 770 1 812
Possible stock drawdown 1 5 131 137
2001/02 Utilization 40 45 1 989 2 074
Food use 40 32 1 761 1 833
of which: local purchase requirement - - 16 16
Non-food use - 12 228 240
Exports or re-exports - 1 - 1
Possible stock build up - - - -
2001/02 Import Requirement 37 - 88 125
Anticipated commercial imports 35 - 83 118
Food aid needs 2 - 5 7
Current Aid Position
Food aid pledges 2 - 5 7
of which: delivered 2 - 5 7
Donor-financed purchases - - 16 16
of which: for local use - - 16 16
for export - - - -
Estimated Per Caput Consumption (kg/year) 3 2 132 137
2001 Production as % of average: 85
2001/02 Import requirement as % of average: 98
FAO/GIEWS - April 2002
FAO/GIEWS: Africa Report No.2 - August 2002 p.41
CEREAL SUPPLY/DEMAND BALANCE FOR THE 2002/03 MARKETING YEAR (in thousand tonnes)
Wheat Rice Coarse grains Total
Previous five years average production 2 82 2 065 2 149
Previous five years average imports 40 1 86 127
2002/03 Domestic Availability 2 63 1 597 1 662
2002 Production (rice in paddy terms) 2 95 1 597 1 694
2002 Production (rice in milled terms) 2 62 1 597 1 661
Possible stock drawdown - 1 - 1
2002/03 Utilization 50 67 2 055 2 172
Food use 50 55 1 776 1 881
of which: local purchase requirement - - 3 3
Non-food use - 12 274 286
Exports or re-exports - - - -
Possible stock build up - - 5 5
2002/03 Import Requirement 48 4 458 510
Anticipated commercial imports 48 4 250 302
Food aid needs - - 208 208
Current Aid Position
Food aid pledges - - 14 14
of which: delivered - - 14 14
Donor-financed purchases - - 3 3
of which: for local use - - 3 3
for export - - - -
Estimated Per Caput Consumption (kg/year) 4 5 155 164
2002 Production as % of average: 79
2002/03 Import requirement as % of average: 402