Targeting aspects of commodity choice in emergency operations are discussed further in Chapter

2, section 2.2.d.

Finally, the District Education Officer was interviewed about the targeting impact of 

School

Feeding

 in Garissa. (The District has a regular School Feeding Program but was not targeted for

the expanded program under the drought EMOP).  His comments agreed with the findings among

pastoralists in Karamoja, Uganda (see Appendix 2.2.a.)   i.e. that school enrolment was low

(perhaps 30% of the age group), but that children from poorer families were at least as likely to be

in school as their richer contemporaries.

3.3.

RECOMMENDATIONS FOR IMPROVED TARGETING

In addition to the recommendations made above about strengthening the linkage of drought

monitoring information to decision makers and standardizing nutritional surveys,  the following

measures are suggested to improve targeting in Kenya.

As noted above, the weak link in Kenya's food security information systems lies in the 

use 

of data

and information rather than their availability.  The unevenness and heterogeneity  of information

from multiple sources (see section 3.1.c.)  also makes it difficult to gain an objective comparative

overview of relative food aid needs in different parts of the country.  During the 1997/8 drought

relief, this led to a perception that areas covered by NGOs strong in collecting credible food

security information and directing it effectively to decision makers (particularly Oxfam)  received

more than a  fair  share of international food aid, i.e. that the Stage 1 targeting was uneven.  It is

suggested that a constructive way to address this problem is to encourage other NGOs to be

similarly pro active in the use of monitoring information from their own areas when food crises

threaten, and to establish a capacity at national level to systematically screen and compare

information from different sources so as to provide as objective a ranking of area needs as

possible.  This role is to some extent played by FEWS, which collates information from all relevant

sources.  However, it is also 

RECOMMENDED that priority be given to establishing  the

25

mandate and capacity to screen and compare information from different areas within the

proposed Drought Management Secretariat in the government system

.  This Secretariat

would then be the central screening unit for assessing slow onset disasters such as drought 

related food crises.   A similar role for sudden onset disasters (such as floods) could appropriately

be established in the NOC.

It is also 

RECOMMENDED that the DMS  together with counterpart/ partner organizations in

26

the international food security community should develop a national level `RAT' (rapid

assessment team) capacity

 with a similar purpose to that at District level   i.e. to follow up early

warning signals from at risk areas in order to assess and quantify relief or other response needs.

An agreed analytical framework and procedure for joint assessment missions co ordinated in this

way could form the basis for a stronger objective comparison of different areas at the planning

stage of food aid operations.  National RATs could also, when needed, assess areas not covered

by the regular DPIRP / ALRMP drought monitoring system.

One specific data gap which hampers assessment of targeting is the lack of comprehensive and

accessible records on international food aid distributions under various programs.  It is therefore

RECOMMENDED that WFP Kenya (possibly under its new VAM unit), and/or FEWS,

establish a data base system for tracking deliveries of food aid to as small a level of

27

disaggregation as practical, and making such data easily and quickly accessible for

analysis.

Targeting food aid to pastoralists is an important issue in all three countries which is outlined in

general terms in Chapter 2.  In Kenya (and in Tanzania)  Oxfam have a policy of starting food aid

A 47