IN

VII. SOCIAL NEEDS AND CONSIDERATIONS

VII.4. CHANNELING AID

It appears that both food and development aid to Afghanistan's rural sector will be better-targeted and more effective to the extent that aid interventions and development projects are formulated with a good understanding of the social, ecological, and economic dynamics of Afghanistan's various agro-ecosystems. WFP's analysis of social groups at the village level and the livelihoods analysis approach advocated by the Afghanistan Research and Evaluation Unit both bring valuable insights to the aid planning process, but, in addition, further inter-agency cooperation is needed.

Sue Lautze et al. (May 2002) recommended that rehabilitation of annual crop agriculture in Afghanistan be deemed a lower priority than (1) securing adequate water supply for all villages in drought-stricken areas, and (2) watering livestock and irrigating orchards and other tree crops (i.e., conservation of assets). The specter of a wealthy farmer using a privately-owned tube well to irrigate his own fields while at the same time causing the water table to sink below the level of shallower wells used for domestic drinking water in nearby villages is a poignant example of a water resources problem that will be difficult to solve.

The same report (Lautze et al. May 2002) recommended that Cash-for-Work (CfW) be used more often, and reliance on Food for Work reduced, so as to protect wheat prices from further decline and to give rural households more flexibility in managing their affairs. The report also recommended making maize more available, as it is a traditional "poverty food" but recently has not been particularly available. Enabling wheat prices to drift upwards would benefit wheat farmers who (1) produce wheat to sell on the market or (2) as tenants (or sharecroppers) are obligated to deliver a certain amount of market value in wheat to the landowner.

Accurate assessment is crucial for aid/relief agencies to be able to formulate and implement the most appropriate policies during the key transition period between emergency and rehabilitation. In May 2002, one theme expressed at the "Code of Conduct" workshop in Kabul was that the focus of ICARDA's work in Afghanistan should transition from emergency seed work to support for a national seed industry by the latter part of 2002. At about the same time, the Lautze et al. report indicated that resource insecurity (and presumably the social instability that it implies) was endemic in all regions of Afghanistan that had been studied, and was increasing. According to the report, U.N. agencies may have under-estimated levels of food insecurity in 2001. What was needed, the report concluded, was a multi-year strategy of expanded relief and development assistance (http://famine.tufts.edu/download/pdf/cash_famine.pdf).

VII.5. SHORT- AND LONG-TERM AID INTEGRATION

Food aid requirements need to be assessed, and delivery implemented, with greater sensitivity to fostering food security in both the short- and longer-terms. Food aid should not be withheld or down-sized on the assumption that livestock assets estimated to be on the target landscape must be converted to "cereal equivalents" for consumption within that season's household food supply.

In fairness to WFP, their Afghan food assessment methodology appears to have evolved considerably between Summer-Fall 2001 and Spring 2002. Any criticism of their earlier methodology may therefore be "no longer needed." As the primary goal of food aid is to "save lives", it is understandable that "sustainability of agriculture issues" would not have been at the forefront of WFP's concerns in Afghanistan in 2001. But, it may also be true that the issue of "livestock assets" and the issue of assumed food reserves caused WFP to under-estimate actual food needs in 2001. "…it is likely that the UN under-estimated food insecurity last year in both levels of assistance it appealed for, as well as the determination of areas that could be classified as extremely vulnerable"(Lautze et al. May 2002. Page 50).

The food security assessment methodology used by WFP Afghanistan in 2001 is described in a WFP publication (WFP Afghanistan, October 2001). According to this methodology, WFP's nutrition team assumed "the minimum per capita food need is calculated on the basis of 2,100 Kcal per person per day." They assumed that 100 g wheat is equivalent to 330 Kcal and therefore the "minimum energy requirement per capita per year" would be:

0.636 kg wheat x 30 days x 12 months = 229 kg wheat per capita per year.

This reasoning would be solid but for the fact that the 229 kg figure never appears in print. Instead, the authors of the October 2001 vulnerability analysis report chose to assume that 20% of food supply would be locally-provided in rural areas in the background, so to speak. In other words, it would be coming from somewhere, but not from farm production or remittances or off-farm income. The key sentence laying out this assumption is: "Assuming people cover 20% of their basic food need, this minimum per capita food need is equivalent to 183 kg per person for twelve months" (WFP. Oct. 2001. p.6).

Estimated farm production, remittances and other income were then all deducted against the residual 183 kg figure in order to calculate net food aid likely to be needed. Nowhere in this publication is any explanation given for how rural communities will "cover" this 20% of their basic food need or whether this "assumed reserve" is unique to rural Afghanistan or also applies to urban areas or is an "across the board" assumption that is applied in all countries needing food aid.

By Spring 2002, this "20% assumed reserve" for Afghanistan's rural areas has more or less disappeared from WFP rapid assessment reports carried out at the district level. One exception is the report on a rapid assessment conducted in Obeh District of Herat Province in March/April 2002:

This seems to be fuzzy logic--to argue that fish that have not been seen, and wild plant food that has not been quantified, represent 20% of the required minimum food supply of poorer groups in a district.

The report on the WFP food security assessment for Tiwara district of Ghor Province is the only report accessed which provides detail on the kinds of wild food collected. Given that a farmer with a sickle can cut about 2 kilograms of grass in about 2 minutes, the statement that a farmer would need "one day" to collect 2 kilograms of wild foods suggests that it is extremely scarce and not a reliable food source for a community already at risk. Net energy yield from this day-long activity could be very low indeed.

Three other reports on district-level rapid food security assessments conducted in the first half of 2002 either do not mention, or barely mention, the role of wild foods. These reports make no mention of an assumed 20% reserve. These rapid assessments are often conducted in remote areas thought to be highly food-insecure. It is not clear whether WFP continues to think that a 20% food reserve assumption makes sense in more central parts of the country and has simply waived this assumption for some highly drought-affected areas. In any case, it would appear that WFP Afghanistan does not yet have substantive figures on which to base an estimate of the amount of annual food requirement being met by wild foods in various agro-eco-niches of the country.

Mission travel in six relatively central Afghan provinces in May revealed that village housing is often built on arid wasteland beside fertile irrigated agricultural land. Other than some birds, the landscape adjoining these villages appeared to be devoid of wildlife. Natural vegetation beyond irrigated areas tended to be exceedingly sparse. Several irrigated farmers interviewed at Bamiyan Center admitted that they had never hiked to the top of the large hills which rose up behind their fields. With the exception of shepherds, some fuel-gathering, and migration to cities for safety or employment, their existence within the confines of the irrigated areas seemed to be remarkably bounded by the extreme aridity of the non-irrigated areas. While it may be true that villagers in particularly remote areas may have more of a hunting and gathering tradition, it would appear that WFP's assumption, published in October 2001, of "20 percent" of food supply covered by non-farm and non-market sources was unwarranted and therefore quietly dropped from subsequent analyses. This assumption may have caused WFP to under-estimate national food requirements by as much as 20% during Fall 2001.

Assumptions about population growth: On the other hand, WFP food security assessment reports assume that district and province population levels have increased uniformly at a rate of 3% per annum since 1991. This assumption is equivalent to an increase over 1991 population of 34% by 2001 and 38% by 2002. In view of mortality caused by civil war and reductions in population due to out-migration to other countries, the 3% population growth per annum may be too high.

Conversion of livestock into "cereal equivalents": In its Fall 2002 report on food vulnerability, WFP assumed that all livestock in a village or district are to be converted into "cereal equivalents" and this value then deducted from estimated food deficits in calculating how much food aid for which an area can qualify.

There are a number of problems with this approach, and it is remarkable that WFP would have adopted such an unrealistic approach to food aid calculations.

The first problem with treating livestock as "liquid assets" is that they are not owned evenly in a village or district. It is not a sound assumption that livestock owners will donate their animals to others who are quietly starving to death. The mission visited villages in Ghazni Province where more than one ethnic group was represented, and in which land and livestock ownership was highly skewed. One landowner in Wardak Province who owned most of the land in the village, told the mission that when a particular tenant household was repeatedly unable to repay debts, they were simply replaced with new tenants. In other words, no communal safety net was operative in that village.

Unlike some NGOs that assess food security using household interviews in Afghanistan, WFP Afghanistan assesses food security by using group estimates of total food assets owned within a village.

Understandably, WFP does not want to be in the position of "giving food stamps to someone with a Mercedes parked in their garage"(analogy of Dr. Michael Latham of Cornell University's Department of Nutrition). But, methodologically, WFP has approached the livestock issue in a way is that is bound to deprive high-risk households of adequate food relief because "there is a Mercedes parked somewhere in the village or the district."

A second problem with this "livestock as liquid assets" assumption is that even if a household or village liquidates all its livestock in a given year, that household or village would be a lot more food-insecure in the following year when there is no longer access to animal traction or farm income from livestock. When viewed beyond the one-year timeframe, the humane and most cost-effective approach to food aid is to respect and reinforce the need for Afghan communities to rebuild their greatly-diminished livestock holdings as one component of the shortest path to regaining food self-sufficiency.

A third problem relates to the approach WFP took to the valuation of livestock assets in Fall 2001. Enumerators asked for information on livestock prices for a three-month period. The average of these time-series data was then used to generate prices for livestock assets. This gross value was then converted to "cereal equivalents" as if all livestock owners could sell all their livestock at these prices and then buy wheat or price-equivalent staples with the proceeds.

Current livestock prices sampled by WFP presumably reflect a market in which a relatively small proportion of the standing herd is coming to market in a given week and there are ample numbers of buyers and sellers. Any wholesale conversion of a village's or district's livestock to "cereal equivalents" would likely trigger a collapse in market prices. And if a livestock owner decided to eat his livestock instead, the household would be eating a normally higher-priced food asset and therefore be likely to experience a caloric shortfall. In various interviews with Kutchi pastoralists, the mission learned that during periods of drought, livestock prices dropped to a fraction of what they had been.

Fourthly, in an effort to provide more transparent methodology and findings to its readership, WFP Afghanistan might consider providing pie charts that show clearly what portions of food requirement are assumed to be coming from where. A non-specialist reading the legend for WFP Afghanistan's food security map of October 2001 might be inclined to think that "more than 100% of food requirement (more than 12 months)" means that such a province has produced a food surplus in the year in question. A careful reading of the WFP methods section reveals that actual food production in that year might be a relatively small part of the assets assumed to be involved in meeting food requirements for that year.

The net effect of these assumptions made in 2001 may be that actual food supply was over-estimated and delivery of food aid to food insecure areas was inadequate. The resultant asset-stripping by insecure households may have resulted in difficult-to-reverse long-term impoverishment that could have been avoided if more generous amounts of food aid had been made available in the first place.

VII.6. ASSESSMENT OF FUTURE AID PROGRAMS TO ENSURE COVERAGE

Future aid programs, such as the program that delivered USAID-funded "seed and fertilizer aid" to largely irrigated farm households in mostly central provinces of Afghanistan in Spring 2002, should be carefully assessed before launch. Seed and fertilizer aid might better be part of a broader plan to provide "integrated agro-economic assistance to poor rural households" across the entire agro-pastoral spectrum at a nationwide level.

Such assistance-especially if pastoralists and landless rural households are involved-may not always include distribution of seed or fertilizers, but-building on the experience of the Spring 2002 fertilizer voucher program-might better consist of a "flexible basket of vouchers cum grant funding" that could paid to qualifying poor rural households across the entire agro-pastoral spectrum in mixes deemed by administering NGOs to be appropriate to each household situation. Relying on careful coordination with WFP/FAO food security assessment survey findings, highest priority in aid targeting would be delivering a flexible package of agro-economic assistance to the most food-insecure communities and households. Many Kutchi pastoral and agro-pastoral communities are thought to have very significant food insecurity; they should also benefit from delivery of aid to the rural sector.