National Guidelines

on

Targeting Relief Food Assistance

Addis Ababa
August 2011

Federal Democratic Republic of Ethiopia

Ministry of Agriculture

Disaster Risk Management and Food

Security Sector
Foreword

This document, which is the outcome of a thorough process of review and stakeholder consultation,
replaces the previous National Food Aid Targeting Guidelines as the major statement of government
guidance on how relief food assistance should be targeted in Ethiopia. It is more than ten years since
those first Targeting Guidelines were issued in 2000. Much experience has been gained and much has
changed since then, and a revision was overdue.
The Disaster Risk Management and Food Security Sector (DRMFSS), which was re-organised under its
new title in 2008, took on the task in 2009. After two years of work I am proud to present such a
comprehensive and focused document.
These new guidelines build on past experience, looking to the future. They are a key step in
implementing the new holistic approach of DRMFSS represented by the shift in national policy from a
focus on reactive relief to a complete disaster risk management cycle inclusive of disaster risk reduction
as well as response, recovery and rehabilitation. They set out the core principles and standards which
should govern the targeting of relief food assistance, in line with international humanitarian standards.
They clarify the proper focus of the disaster response system on temporary, transitory-acute needs rather
than chronic food insecurity. They stress the fundamental principles that targeting should be based on
need, and should prioritise the most vulnerable and the most urgent cases.
Guidelines are a tool for implementation. This document should therefore be taken as a starting point
for renewed efforts by all concerned agencies to work together in ensuring effective and accountable
targeting of relief. It should serve as a reference and a resource for both government staff and
humanitarian partners.
Well-targeted relief assistance will not only safeguard the lives and livelihoods of disaster-affected
people, but will also safeguard the overall economic growth of the country and reduce future
dependency, by enabling people to recover their self-reliant livelihoods as quickly as possible after a
shock. These guidelines are a first fundamental step in this direction.

Ato Mitiku Kassa

State Minister
Disaster Risk Management and Food Security Sector
Ministry of Agriculture

i
Contents
Acronyms ................................................................................................................................................... vi

PART ONE: INTRODUCTION ........................................................................................................... 1

1. The importance of targeting ..........................................................................................................1
2. The need for new guidelines ..........................................................................................................2
2.1. Changes since the existing guidelines were issued ....................................................................2
2.2. Directions of change for the future ...........................................................................................2
3. Scope and purpose of the guidelines .............................................................................................4
3.1. Scope ..........................................................................................................................................4
3.2. Intended uses and users ............................................................................................................4
3.3. Interface with other guidelines ..................................................................................................5
3.4. Implementation requirements...................................................................................................6
4. Organisation of the contents .........................................................................................................6

PART TWO: CORE PRINCIPLES ........................................................................................................ 7

5. Targeting relief as a part of Disaster Risk Management ................................................................7
6. A multi-hazard, multi-sectoral framework .....................................................................................9
6.1. Hazard types ............................................................................................................................10
6.2. Disaster types...........................................................................................................................12
6.3. Multi-sectoral response ...........................................................................................................12
7. Transitory-acute versus chronic food insecurity: defining the target group for relief .................13
8. International minimum standards for relief targeting .................................................................18
9. Guiding principles of the DRM Policy applied to targeting ..........................................................18

PART THREE: THE TARGETING TOOLBOX......................................................................................... 21

10. Six key questions of targeting ......................................................................................................21
11. Levels of targeting ........................................................................................................................22
12. Choosing targeting criteria ...........................................................................................................23
12.1. Linking targeting criteria with vulnerability and needs analysis ..............................................23
12.1.a. Nutritional criteria for prevention or cure of malnutrition .............................................24
12.1.b. Economic criteria for early response and livelihood protection .....................................25
12.1.c. Social and categorical criteria for targeting vulnerable groups .......................................25
12.2. Combining targeting criteria ....................................................................................................26
13. Targeting methods and mechanisms ...........................................................................................26
13.1. Administrative targeting ..........................................................................................................26
13.2. Community targeting ...............................................................................................................26
13.3. Self-targeting............................................................................................................................27
13.4. Institutional targeting ..............................................................................................................28
13.5. Market-based targeting ...........................................................................................................30
13.6. Combining methods .................................................................................................................30
14. Targeting ‘errors’ and problems...................................................................................................30
14.1. Perfect (theoretical) targeting .................................................................................................30
14.2. ‘Benefit incidence’ errors: exclusion and inclusion..................................................................33
14.3. ‘Benefit level’ errors: dilution and leakage ..............................................................................34
14.4. Timing errors ............................................................................................................................34

ii
 14.5. Causes of targeting error .........................................................................................................35
15. Tools for monitoring and evaluation............................................................................................35
15.1. Types of monitoring .................................................................................................................35
15.2. Monitoring the process (how well is the targeting system working?).....................................37
15.3. Monitoring the outcome (who received what?)......................................................................37

PART FOUR: ETHIOPIA’S NATIONAL TARGETING SYSTEM ............................................................... 40

16. Targeting throughout the disaster-response cycle ......................................................................40
17. Targeting levels, institutions and roles ........................................................................................42
17.1. Level 1: Geographical targeting from federal to wereda level ................................................42
17.1.a. Needs assessments ..........................................................................................................42
17.1.b. PSNP and relief needs......................................................................................................43
17.1.c. Information flows ............................................................................................................43
17.1.d. Regional verification ........................................................................................................44
17.1.e. ‘Hot-spot’ Prioritisation ...................................................................................................45
17.2. Level 2: Geographical targeting from wereda to community ..................................................45
17.3. Level 3: Beneficiary targeting...................................................................................................46
18. The Household Economy Approach applied to targeting ...........................................................46
18.1. Assessing multi-hazard impacts and coping capacity ..............................................................46
18.2. Geographical targeting by livelihood zone ..............................................................................48
18.3. Seasonality: timing and duration of relief needs .....................................................................48
18.4. Household targeting: setting provisional criteria ....................................................................49

PART FIVE: BENEFICIARY TARGETING .......................................................................................... 51

19. Combined community and administrative targeting ...................................................................51
19.1. Situations where community targeting may not be applicable ...............................................52
19.2. Roles of government, partners and community ......................................................................52
20. Process and governance of community-administrative targeting ...............................................54
20.1. Institutional arrangements ......................................................................................................54
20.1.a. General lines of communication and accountability .......................................................54
20.1.b. Wereda and kebele responsibilities ................................................................................54
20.1.c. Community representation .............................................................................................56
20.2. Key steps at community level ..................................................................................................58
20.3. Transparency: information is power ........................................................................................58
20.4. Accountability ..........................................................................................................................59
20.4.a. Correcting errors: appeals and complaints......................................................................59
20.4.b. Replacing bad decision-makers .......................................................................................61
20.4.c. Record-keeping, registration and ration cards ................................................................61
21. Fine-tuning the local targeting criteria.........................................................................................62
21.1. Targeting according to need ....................................................................................................62
21.2. Respecting community values .................................................................................................63
21.3. Avoiding dependency traps .....................................................................................................64
22. Registering women as food entitlement holders.........................................................................65
23. General food distribution (GFD)...................................................................................................65
23.1. Ration rates and the ‘food basket’...........................................................................................66
23.2. Targeting by family size............................................................................................................67

iii
 23.3. Duration of relief eligibility ......................................................................................................67
24. Supplementary foods (SF) ............................................................................................................67
24.1. Nutritional targeting of curative SF (for acutely malnourished individuals) ............................67
24.2. Categorical targeting of preventive SF or ‘relief CSB’ (for vulnerable groups) ........................68
24.3. Co-ordinating GFD with Supplementary and Therapeutic Feeding .........................................68
25. Co-ordinating relief with the Safety Net (for PSNP weredas only) ..............................................69
25.1. Phased response to shocks: PSNP, Risk Financing and relief ...................................................69
25.2. Household targeting of PSNP and relief assistance .................................................................69
26. Unconditionality of relief assistance ............................................................................................71
27. Adapting targeting processes and criteria to local contexts ........................................................72
28. Monitoring for improved targeting ..............................................................................................72

PART SIX: CROSS-CUTTING ISSUES ............................................................................................. 74

29. Gender equity...............................................................................................................................74
30. Child protection............................................................................................................................76
31. HIV/AIDS, chronic illness and disability ........................................................................................76

GLOSSARY ................................................................................................................................. 78
REFERENCES AND FURTHER READING ............................................................................................... 84

iv
LIST OF FIGURES, TABLES AND BOXES

Figure 1: Relief Assistance in the Disaster Risk Management Framework .....................................................8

Figure 2: Scale and frequency of disasters triggered by different hazards in Ethiopia ................................11
Figure 3: Separation of mandates for targeting transitory and chronic food insecurity ..............................14
Figure 4: Effects of shocks on the resources of vulnerable and resilient households ..................................16
Figure 5: Hypothetical situation at needs assessment .................................................................................31
Figure 6: Effect of blanket (untargeted) distribution ....................................................................................31
Figure 7: Effect of perfect targeting..............................................................................................................32
Figure 8: Exclusion and inclusion errors .......................................................................................................33
Figure 9: The chain of actions needed to ensure good targeting ...............................................................40
Figure 10: Overview of the relief targeting system, from federal to beneficiary level...............................41
Figure 11: Methodological Framework: the Household Economy Approach (HEA) .....................................47
Figure 12: Targeting by Livelihood Zone - different outcomes between LZs within one wereda ................49
Figure 13: Targeting by Wealth Group - different outcomes within one livelihood zone ............................50
Figure 14: General lines of accountability and information flow from Region to community .....................54
Figure 15: Lines of appeal: checks and balances ..........................................................................................60

Table 1: Comparison of previous and new targeting guidelines ....................................................................3

Table 2: National Guidelines and Regional Handbooks ..................................................................................5
Table 3: Types of hazard, disaster, and response .........................................................................................10
Table 4: ‘Composite food insecurity’ matrix: combining duration and severity..........................................15
Table 5: Guiding Principles of the DRM Policy Applied to Targeting ............................................................18
Table 6: Types of vulnerability and appropriate targeting criteria ...............................................................24
Table 7: Advantages and disadvantages of different targeting methods .....................................................29
Table 8: Key questions and methods for monitoring relief targeting...........................................................36
Table 9: A basic checklist for monitoring the targeting process ...................................................................38
Table 10: Questions and indicators for monitoring targeting outcomes .....................................................39
Table 11: Types of Coping Strategy in HEA Analysis .....................................................................................48
Table 12: Administrative, partner and community roles in relief targeting .................................................53
Table 13: Recommended membership of wereda and kebele bodies responsible for relief targeting .......55
Table 14: Broad and detailed criteria for household targeting ....................................................................63

Box 1: Example of a disaster requiring a complex multi-sectoral response .................................................13

Box 2: International Standards for Targeting in Disaster Response .............................................................17
Box 3: Characteristics of good targeting criteria ..........................................................................................23
Box 4: Seven key steps in the community-administrative targeting process ...............................................57
Box 5: Standard full food basket for GFD in Ethiopia ...................................................................................66
Box 6: PSNP rules for households affected by shocks ..................................................................................70
Box 7: People who should never be required to work in exchange for relief..............................................71
Box 8: Gendered risks and vulnerabilities in Ethiopia ..................................................................................74
Box 9: Proxy indicators for households affected by HIV/AIDS ......................................................................77

v
 Acronyms

ALNAP Active Learning Network for Accountability and Performance in Humanitarian Action

BPR Business Process Re-engineering

CB Contingency Budget (PSNP)

CCI Complementary Community Infrastructure (Food Security Programme)

CFW Cash for work

CRS Catholic Relief Services

CSB Corn-soy blend

DA Development Agent

DFID Department for International Development (UK)

DPPB Disaster Prevention and Preparedness Bureau (Regional level)

DPPC/ A former Disaster Prevention and Preparedness Commission/ Authority (now superseded by DRMFSS)

DRM Disaster Risk Management (= DRR + response, recovery and rehabilitation)

DRMFSS Disaster Risk Management and Food Security Sector (Federal Ministry of Agriculture)

DRMSPIF Disaster Risk Management Strategic Programme and Investment Framework

DRR Disaster Risk Reduction (= pre-disaster prevention, mitigation and preparedness)

EFSA Emergency Food Security Assessment (WFP)

ENCU Emergency Nutrition Coordination Unit (DRMFSS)

EWR Early Warning and Response

EWRD Early Warning and Response Directorate (DRMFSS)

EWS Early Warning System

FEG Food Economy Group

FFW Food for work

FS Food Security

FSCD Food Security Co-ordination Directorate (DRMFSS)

FSTF Food Security Task Force (in PSNP weredas, kebeles and communities)

HABP Household Asset Building Programme (Food Security Programme)

HEA Household Economy Approach

vi
HEW Health Extension Worker

HFA Hyogo Framework for Action (UNISDR)

JEOP Joint Emergency Operation Plan

LIAS Livelihoods Impact Analysis Sheet

LIU Livelihoods Integration Unit (DRMFSS)

LZ Livelihood Zone

MoA Ministry of Agriculture

MoARD former Ministry of Agriculture and Rural Development (until September 2010)

NGO Non-Governmental Organisation

NPDPM National Policy on Disaster Prevention and Management (1993)

PA Peasant / Pastoralist Association (= rural kebele or sub-district)

PLWHA People living with HIV/AIDS

PSNP Productive Safety Net Programme

RFM Risk Financing Mechanism (PSNP)

SNNPR Southern Nations, Nationalities and Peoples‟ Region

UN United Nations

UNISDR United Nations International Strategy for Disaster Reduction

USAID United States Agency for International Development

WFP UN World Food Programme

vii
TARGETING RELIEF FOOD ASSISTANCE PART ONE: INTRODUCTION

PART ONE: INTRODUCTION

1. The importance of targeting

Targeting – ensuring that assistance reaches the right people, in the right kind and quantity, and at the
right time and place – is at the heart of any disaster response operation. At the same time, targeting in
emergency situations is an inherently challenging process which raises a common set of problems and
dilemmas across diverse contexts, both within Ethiopia and in other countries. A targeting system does
not operate mechanically: it requires constant feedback, judgement and accountability. These guidelines
draw on lessons learned in Ethiopia, and on the international experience of WFP and other humanitarian
agencies, to update the official guidance on targeting of relief food assistance and to reaffirm the
commitment of DRMFSS and its humanitarian partners1 to its continuing improvement.

“The purpose of targeting is to meet the needs of the most vulnerable, while providing aid
efficiently and in a way that minimises dependency.”
Sphere Project 2004:37

The main reasons for targeting (rather than blanket or random distribution) of relief food assistance are:
1. Humanitarian principles. In emergency situations, aid should always be given according to need,
following international standards of impartiality and prioritisation of the most vulnerable. Targeting
by agreed and verifiable criteria, derived from systematic assessments of vulnerability and needs,
supports these principles.
2. Effectiveness. In order to achieve its intended impact on food insecurity and acute malnutrition, it is
obviously essential that relief assistance reaches the right people. Poorly targeted assistance may fail
to meet the core objectives of emergency relief: relieving suffering, saving lives and saving livelihoods.
3. Avoidance of harm. Untargeted or poorly targeted assistance (particularly food aid), spread too
widely and in large quantities, may undermine local production or markets and may encourage
dependency.
4. Resource shortfalls. In emergency operations, it often happens that the resources available for
distribution are less, or later, than had been requested or planned. In these situations it is necessary
to prioritise the people in most urgent need. A transparent, needs-based targeting system provides
an objective basis for doing this.
5. Efficiency. In a wider sense, relief funds and commodities are always scarce and valuable resources.
Targeting helps to make the most economical use of them, to minimise waste, and to achieve the
maximum benefit for disaster-affected people from a given quantity of resources. However, targeting
itself is not necessarily low-cost: resources are needed to manage its implementation and monitoring.
6. Accountability. Government authorities, donors and partner organisations need to know that their
aid actually reaches the people who need it, and that it has not been diverted or wasted. Good
targeting systems ensure this, and at the same time they generate information about where the
resources went and what impact they had on beneficiaries’ lives. Transparent targeting also enhances
the accountability of government and partner agencies to beneficiary communities.

1 Humanitarian partner organisations are UN, donor, non-governmental and civil society organisations (such as the
Ethiopian Red Cross Society) who are participating with DRMFSS in the provision of humanitarian relief
assistance to disaster-affected people.

1
TARGETING RELIEF FOOD ASSISTANCE PART ONE: INTRODUCTION

2. The need for new guidelines

2.1. Changes since the existing guidelines were issued

The previous Food Aid Targeting Guidelines, issued in 2000, would be due for revision in light of the
intervening decade of experience even if nothing had changed in the context of disaster response in
Ethiopia. In fact, there have been significant changes which require a rewriting of the guidelines:
 A paradigm shift has taken place in the government approach to dealing with emergencies, from a
drought-focused disaster response model to a multi-hazard risk management approach. The 1993
National Policy on Disaster Prevention and Management, on which the previous targeting guidelines
were based, has been replaced by a Disaster Risk Management (DRM) framework. The DRM policy
calls for a decentralised and community-based approach to disaster risk management, including relief
activities.
 Since 2005, the PSNP (Productive Safety Net Programme) has been rolled out across most of the
country, with a mandate to replace a large part of the previous relief system with predictable food
and cash transfers for chronically food insecure households. The PSNP has its own targeting
procedures and guidelines. On the one hand, this new initiative has prompted a re-focusing of relief
operations on transitory food insecurity. On the other hand, it raises some practical coordination
issues for relief targeting, particularly in PSNP weredas.
 Over the same period, there have been significant changes in the methodology employed in
geographical targeting from the federal to the wereda level. The Household Economy Approach
(HEA) now forms the analytical basis for the national needs assessments and relief planning figures.
These new guidelines aim to extend the logic of the HEA approach to the beneficiary level, making
clearer links between needs assessment, geographical and household targeting for a more ‘joined-
up’ system.
 In international humanitarian policy, ‘food aid’ has been largely replaced by the wider category of
‘food assistance’, meaning assistance for people to meet their food needs: this may be provided
either in kind or in the form of cash or vouchers. If provided in kind, it may be either locally
purchased or imported.
 Lastly, a wide-reaching programme of Business Process Re-engineering (BPR) has changed the
institutional structures at various levels of government in Ethiopia, necessitating a revision of the
exact lines of responsibility and accountability involved in relief targeting.

Table 1 summarises the key differences between the previous guidelines and this revised document.

2.2. Directions of change for the future

Based on the principles of the DRM Policy, and taking account of the above changes, these guidelines
encourage continuing efforts to improve the following aspects of relief targeting. Several of these
objectives were included in the previous Food Aid Targeting Guidelines but have not yet been fully
achieved, while some are new.
Decentralisation. Responsibility for relief targeting is to be increasingly delegated to the Regions, and
early warning will increasingly rely on wereda information rather than national seasonal assessments.
This gives scope for more local adaptation and tailoring of the targeting process, within the framework of
national rules and principles set out here.
Community empowerment. The new policy re-states the “decisive role” of the community in disaster
risk management activities, including relief targeting. It is recognised that the balance of power in the
‘administrative-community’ targeting process has been variable and sometimes unclear in the past.
These guidelines aim to clarify the respective roles in this process and to support genuine community
participation in decision-making.

2
TARGETING RELIEF FOOD ASSISTANCE PART ONE: INTRODUCTION

Table 1: Comparison of previous and new targeting guidelines

National Food Aid Targeting Guidelines National Guidelines on Targeting Relief

(2000) Food Assistance (2011)
Policy basis National Policy on Disaster Prevention and Disaster Risk Management Policy (2010)
Management (1993)  Multi-hazard framework
 Linking relief and development through  Disaster response part of whole DRM
Employment Generation Schemes (EGS) cycle, not the primary strategy
 Drought-focused  Relief focused on short-term emergency
 Food aid used to address both chronic needs
and acute food insecurity
International not covered Guidelines linked to Sphere Minimum
standards Standards for Disaster Response as well as
Guiding Principles of the DRM Policy.
PSNP (did not exist in 2000) Addresses coordination of relief and PSNP
targeting
Fairness States “fairness is the key word”, but without Clearly defines “fairness” as needs-based
defining it. Open to different interpretations. targeting, based on impartial assessment of
all disaster-affected people and their different
needs and vulnerabilities.
Levels of Operational guidance on beneficiary targeting Main operational focus on beneficiary
targeting only, with an annex on area targeting. targeting.
Includes guidance on targeting throughout the
relief cycle, national targeting system and
geographical targeting.
Linking not covered Promotes „joined-up targeting‟ through logical
geographical connection of household criteria with needs
and beneficiary and vulnerability assessments
targeting Connects beneficiary selection criteria with
the HEA needs assessment framework
Pastoralist Not covered in background studies, though Background consultations included pastoralist
areas pastoralist Regions were consulted on draft. Regions.
Separate guidelines for pastoralist areas National guidelines designed to be relevant to
recommended but not produced. all Livelihood Zones as defined by HEA
baselines, including mainly pastoralist areas.
Regions have flexibility to adapt
implementation to all local livelihoods and
social structures.
Regional States that Regions can adapt guidelines to Sets out framework for Regional adaptations,
adaptations local conditions, but no details. specifies elements of the guidelines that need
local detail.
Supplementary not covered Explains principles and target groups for
Foods supplementary feeding.
Gender equity not covered
Addressed as cross-cutting issues linked to
Child protection not covered
specific guidance points.
HIV/AIDS not covered

3
TARGETING RELIEF FOOD ASSISTANCE PART ONE: INTRODUCTION

Enhanced partnerships between government and humanitarian organisations at all levels of the targeting
system are essential, in order to make best use of the available skills and capacities on both sides, in the
interests of Ethiopia’s disaster-affected people.
‘Joined-up targeting’. Clearer logical links between the needs assessment findings, geographical
targeting and household targeting criteria will enhance the transparency of resource allocations and
facilitate awareness-raising about the aims and limitations of relief assistance. The analytical framework
of HEA offers a basis for improvement in this direction.
Accountability and monitoring. Establishing effective appeals systems for the timely correction of
exclusion and inclusion errors has proved challenging at community level. Continued efforts are needed
in this direction, and greater participation by humanitarian partners at this level could help to strengthen
checks and balances. Accountability also requires systematic record-keeping and reporting, and regular
corrective monitoring of the targeting process and outcomes.
Transparency and information sharing. All of the above improvements require transparency, including
regular and timely information exchange among all partners in relief targeting, at all levels of the system.

3. Scope and purpose of the guidelines

3.1. Scope
The guidelines cover the principles and practice of targeting relief food assistance, within a wider
framework of multi-hazard disaster management and multi-sector response. Food assistance
(particularly in-kind food aid) is not always the appropriate, or the only, response to a given emergency:
non-food needs are also critical. A first step in targeting is to decide what kinds of assistance are needed.
However, these guidelines do not specifically address the targeting of non-food assistance. Although
most of the core principles and many of the procedures set out here will also be applicable to non-food
assistance, key elements including the definition of the target group and the suggested targeting criteria
relate specifically to food needs. Targeting relief assistance in other sectors (such as water, health and
education) requires specialist guidance from the relevant line ministries.
It is intended that these guidelines should apply to rapid-onset as well as slow-onset disasters, to urban
as well as rural food crises, and to innovative modalities of food assistance (such as cash transfers,
vouchers, and potentially market-based assistance in urban areas) as well as in-kind food aid. However, it
must be acknowledged that the Ethiopian relief system has so far focused more on slow-onset rural crises
and on distributing food commodities, and that this is reflected in the balance of the guidelines. It is
therefore recommended that further detailed targeting guidance should be developed for rapid-onset
disasters, urban food crises and new modes of food assistance as greater experience is gained with each
of these. Such additional guidance could either be added to the next revision of the main guidelines, or
issued as a supplement.
The guidelines aim to be normative, that is to set out how relief food assistance should be targeted,
rather than simply describing how the Ethiopian system currently works. At the same time, they aim to
be pragmatic, taking account of existing capacities and structures, building on what is already in place
and considering what is likely to be feasible. If we simply base the guidelines on established practice,
nothing will change. At the other extreme, if we advocate a perfect system that is impractical in current
conditions, nothing will be implemented.

3.2. Intended uses and users

The first purpose of the guidelines is to provide a clear and agreed basis for the continuing improvement
of targeting practice in the Ethiopian disaster response system. All agencies involved in distributing relief
food assistance should adhere to the principles and rules they contain. A degree of flexibility is allowed
to adapt rules and procedures to diverse local conditions, provided the principles are observed.

4
TARGETING RELIEF FOOD ASSISTANCE PART ONE: INTRODUCTION

Secondarily, the guidelines are intended for use in the design of training and capacity development
programmes, and as a reference for anyone working in related fields.
The guidelines are written primarily for decision-makers, technical staff, programme designers and
managers in government and humanitarian partner organisations at national and regional levels. They
may also be a useful reference for government and partner staff implementing relief distributions or
supporting capacity development at wereda and community levels, although shorter Targeting
Implementation Handbooks (containing the core principles and practical instructions) are also planned.

3.3. Interface with other guidelines

These guidelines are one of a set of technical guidance documents being produced or updated by
DRMFSS. Since the targeting of relief food assistance is interlinked with many aspects of disaster risk
management and food security, there are a number of points in the text where the reader is referred to
other specialist guidelines. The targeting guidelines are designed to be consistent with these other key
documents, and not to replace or contradict any of them. References to other official guidelines are
highlighted in the text with a book symbol, like this:

 See DRMFSS 2010c, Early Warning and Response Guidelines (forthcoming 2011)

All documents referred to are also listed alphabetically by author under ‘References and Further
Reading’.

Table 2: National Guidelines and Regional Handbooks

Issued by Purposes Primary audience

National Federal  Clarify government policy, principles  Government and humanitarian partner
Targeting DRMFSS and standards relating to targeting of organisations at national and regional
Guidelines relief food assistance levels
 Support accountability and shared  Professional and technical staff;
understanding among partners decision-makers
 Provide technical information and
reference material on relevant
targeting methods
 Set the general rules and criteria for
relief targeting, in line with national
policy

Targeting Regional  Communicate core content of  Grassroots implementers

Implementation authorities guideline to local-level implementers
Handbooks  Wereda and kebele officials
 Provide practical, accessible
instruction on how to target relief  Field staff of government and partner
beneficiaries organisations

 Adapt details (e.g. eligibility criteria  Community members and

and community processes) to local representatives
realities
 Specify institutional responsibilities
and lines of accountability for targeting
within the Region

5
TARGETING RELIEF FOOD ASSISTANCE PART ONE: INTRODUCTION

3.4. Implementation requirements

The major responsibility for implementing these guidelines rests with the Regions. It is expected that
each Region will translate the core principles and general rules set out in the guidelines, decide the
detailed delegation of responsibilities and lines of accountability within the Region (according to the BPR
and decentralisation processes, and in consultation with relevant partners), and produce an adapted
Implementation Handbook on beneficiary targeting for use from wereda to community level. The
handbooks should take account of the diversity in social, cultural and livelihood contexts in each Region.
The complementary purposes of these guidelines and the planned handbooks are summarised in Table 2.
DRMFSS will support and coordinate the development and piloting of the handbooks, and will welcome
the participation of humanitarian partners in this process.
These guidelines should be reviewed and if necessary revised to ensure that they are correct, up to date
and useful, ideally after the first year of implementation and periodically thereafter.

4. Organisation of the contents

The guidelines are divided into six Parts as outlined below. Each Part is divided into short numbered
sections on particular topics, and the table of contents is designed to function as a ‘menu’ so that users
can easily find the sections they need. While the guidelines are structured as a coherent whole which can
be read from beginning to end, it is also possible to read each section separately for quick reference or
training.
Part One is this Introduction.
Part Two sets out the Core Principles underlying the whole of the guidelines, in relation to current
Ethiopian government policy and international standards. It situates the targeting of relief food
assistance in the context of the multi-hazard Disaster Risk Management paradigm, and defines the target
group for relief food assistance.
Part Three provides a technical Toolbox of targeting methods and terminology, which can be used as a
reference or training resource and which explains the concepts relevant to the more operational content
of Parts Four and Five.
Part Four gives an overview of Ethiopia’s National Targeting System, including the needs assessment
process, geographical targeting from federal to wereda level and from wereda to kebele, and the
application of the Household Economy Approach to targeting.
Part Five contains more detailed guidance and rules on Beneficiary Targeting, meaning the procedures
and criteria for identifying eligible households or individuals within disaster-affected communities. This is
the core of the guidelines, and replaces the previous Food Aid Targeting Guidelines issued in 2000 as the
main statement of DRMFSS rules on targeting at this level. Harmonisation of beneficiary targeting
between relief food assistance and the PSNP is dealt with in this part of the guidelines.
Finally, Part Six focuses on three major Cross-cutting Issues – gender equity, child protection and chronic
illnesses including HIV/AIDS– and draws together guidance points on targeting in relation to each of these
issues.
Key terms are listed alphabetically, with definitions and explanations, in the Glossary at the end of the
guidelines.

6
TARGETING RELIEF FOOD ASSISTANCE PART TWO: CORE PRINCIPLES

PART TWO: CORE PRINCIPLES

5. Targeting relief as a part of Disaster Risk Management

One of the major changes that has taken place since the previous Targeting Guidelines were issued in
2000 is the ‘paradigm shift’ in national policy from a focus on reactive relief (that is, responding to
emergencies when they happen) to the more holistic approach of Disaster Risk Management (DRM).2
The DRM framework has six pillars: prevention, mitigation and preparedness (which constitute ex-ante
disaster risk reduction), plus disaster response, early recovery and rehabilitation. Thus, disaster
response remains a key component of the overall policy, but it is no longer the primary strategy. This
section examines the implications of this and related changes for the targeting of relief food assistance.

The second of the DRM Policy’s four specific objectives is to “save lives and protect livelihoods in the
event of disasters and ensure the recovery and rehabilitation of all disaster-affected
populations.”
(FDRE 2010a)

The previous policy basis for emergency relief was contained in the National Policy on Disaster
Prevention and Management (NPDPM, TGE 1993), which was co-ordinated at federal level by the Disaster
Prevention and Preparedness Commission (DPPC, 1995-2004) and then the Disaster Prevention and
Preparedness Agency (DPPA, 2004-7). Although these policy and institutional mandates did in theory (as
the names suggest) include disaster prevention, preparedness, and management, their effectiveness in
these areas was hampered by a lack of resources and by the constant demands of the annual food aid
cycle, which until 2005 included aid to millions of chronically food insecure people who are now
supported by the Productive Safety Net Programme (PSNP).
The DRM policy identifies a number of shortcomings in the design of the NPDPM, particularly its relative
neglect of non-food relief needs and its single-hazard focus (dealing primarily with slow-onset, drought-
induced crises in food production, to the neglect of other hazards – see Section 6.1 - and urban or fast-
onset disasters). It also notes a number of implementation challenges relating to targeting that were
encountered under the previous relief system:
“Targeting of relief to the destitute in some cases encouraged the depletion of productive
assets, rendering beneficiaries dependent on cycles of relief.
Errors of exclusion in distribution resulted from inadequate resources, quota requirements,
geographical remoteness, and other factors.
Errors of inclusion resulted from inflated estimates of beneficiaries and the frequent
distribution of relief resources to the same areas and other factors.” (FDRE 2010a)
The first of these points expresses a classic welfare-dependency trap, which is addressed in Part Five. The
remaining two points highlight the dilemma that poor targeting can result from either under-estimating
or over-estimating needs: therefore, effective targeting must begin from accurate needs assessment.
Later sections of the guidelines come back to this, explaining the importance of seeing good targeting as

2 At the time of writing, the new policy document is in its 10th draft version and awaiting ratification (FDRE 2010a).
Many elements of the approach are already being applied by DRMFSS. The government is also preparing a DRM
Strategic Programme and Investment Framework (DRMSPIF) which will serve as guidance for the implementation
of the Policy by various actors, through a range of programme components, in the next five years (2011-2015).

7
TARGETING RELIEF FOOD ASSISTANCE PART TWO: CORE PRINCIPLES

an outcome which depends on a number of interconnected actions at different stages of the disaster-
response cycle.
The vision of the Ethiopian government and its international humanitarian partners is that the need for
relief, especially food aid, will be diminished over time by effective disaster risk management (reducing
the frequency and impact of disasters) and by general growth and development (increasing the resilience
and coping capacity of the population when disasters do happen). At the same time, the government’s
Growth and Transformation Plan for 2011-2015 (FDRE 2010b) recognises that disasters, particularly
climatic hazards affecting agriculture3, are a potential threat to development gains both at household and
national level, and that effective risk management and disaster response are therefore essential to
safeguard the sustainability of economic growth. Targeting, including temporal targeting (that is,
timeliness in beginning and ending distributions), is important to the effectiveness of a relief operation
regardless of its size.

Figure 1: Relief Assistance in the Disaster Risk Management Framework

Figure 1 represents the place of relief food assistance within the broader framework of Ethiopia’s new
DRM Policy and projected growth path. It shows DRR (disaster risk reduction) activities as an on-going
thread during pre- and post-disaster phases, with the PSNP (shown in green) underpinning development
investments by providing predictable food and income transfers during the hungry season to chronically
food insecure households in selected vulnerable areas. When a disaster event strikes (or a slow-onset
crisis reaches a tipping point), the first response in PSNP weredas is to use the regional contingency fund
to meet additional transitory needs. If this is not sufficient, then the Risk Financing Mechanism is
activated to scale up the safety net temporarily. If the emergency needs exceed the Risk Financing
ceiling, and/or if the crisis affects areas not supported by the PSNP, then an emergency relief operation is
needed: this may include both food and non-food relief (such as health, water, education, and various
livelihood support interventions), as discussed below. The relief operation must be targeted according
to an impartial needs assessment; and it must be timed correctly to provide relief during the period of
need, phasing out at the right time to support early recovery from the crisis, and ensuring that livelihoods

3 The term „agriculture‟ throughout these guidelines includes all crop and livestock production activities.

8
TARGETING RELIEF FOOD ASSISTANCE PART TWO: CORE PRINCIPLES

are protected and restored. The appropriate time to phase out relief distributions should be assessed in
each specific situation, to ensure people not only survive but also recover. If these conditions are met, as
the figure illustrates, the emergency response can protect and support livelihoods, minimise disaster
impacts, and ensure a timely return to the expected upward path of growth and development.
Capacity development during the pre-disaster phase should include the strengthening of knowledge and
procedures for targeting relief food assistance, so that targeting can be implemented in a timely and
effective manner when needed. A major initiative in this direction is the Wereda Disaster Risk Profiling
Programme (see DRMFSS 2009).
To sum up, targeting is central to achieving effective emergency relief within the overall framework of
disaster risk management. Proper targeting supports the objectives of both disaster risk management
and poverty reduction, by ensuring that people affected by a disaster can recover quickly and do not
suffer irreversible harm to their health, growth (in the case of children), or livelihoods. As the DRM
policy states,
“Every effort will be made to ensure timeliness, adequacy, rationality, and appropriateness
of disaster response measures in order to mitigate effectively disaster-related loss of lives
and livelihoods.......
 Disaster response strategies shall protect the dignity and basic human rights of affected
populations.
 The protection of the rights and livelihood systems must be of equal importance to the
provision of goods and services.
 The nature and scope of disaster response measures shall be based on the actual needs
of at risk populations.” (FDRE 2010a, emphasis added)
Without such support there is a danger that disaster impacts will not only threaten lives but also
undermine any gains in growth and food security. Proper targeting also limits the distribution of relief
assistance, directing it only to the people, places and periods where it is needed, thus avoiding
disincentive effects for local food producers and potential dependency among beneficiaries. These
principles apply to both fast and slow-onset disasters, as the following sections explain.

6. A multi-hazard, multi-sectoral framework

The new DRM paradigm, in line with the international disaster risk reduction framework developed by
the UN 4, starts from a multi-hazard analysis of disaster risks. This considers the whole range of natural
or human-induced hazards which may cause disasters within the country. The potential of a hazard to
cause a disaster depends on both the severity of the event, and the vulnerability of the population to the
impacts of that event. A single hazard or a combination of hazards may cause different types of disaster,
which are usually classified in the emergency food security literature as fast-onset, slow-onset or
complex.
For effective targeting, it is important to start from a clear logical analysis of the type of hazard involved
and its impacts, the vulnerabilities of the population affected, the type and expected trajectory of the
disaster situation, and therefore the type(s) of relief needed. Table 3 gives some examples to clarify the
relationships between hazards, disasters and response options.

4 UNISDR (2007). Hyogo Framework for Action 2005 - 2015: Building the Resilience of Nations and Communities to
Disasters, United Nations International Strategy for Disaster Reduction. www.unisdr.org/hfa UNISDR (2007).

9
TARGETING RELIEF FOOD ASSISTANCE PART TWO: CORE PRINCIPLES

Table 3: Types of hazard, disaster, and response

Different types of different types of different types of

can trigger requiring
HAZARD... DISASTER... RESPONSE.
e.g. e.g. e.g.
Natural:
Fast-onset disasters Food assistance
drought
caused by sudden (in-kind or cash/ voucher)
flood
calamities such as
human diseases Non-food basic items (e.g.
earthquakes, floods,
livestock diseases cooking utensils, blankets)
violence, etc. Impacts
crop pests or diseases
may include loss of food Health and Nutrition
earthquakes
stocks, crops and
volcanic eruptions WASH (Water, Sanitation
livestock; disruption of
climate change and Hygiene)
markets and income
sources; injury and Livelihood support (assets,
Human-induced: disease; displacement or inputs, markets etc)
conflict loss of homes.
war Education
Slow-onset disasters
transport accident Shelter
caused by gradual events
fire (forest or urban)
such as drought, or by Infrastructure (transport,
pollution
cumulative shocks and communications)
environmental damage 5
stresses that erode
market or price shocks Protection / security
people‟s assets and
trade disruption
coping capacity until they
are unable to meet their
food and other basic
needs.
Complex emergencies
in which natural or
human-made hazards are
combined with conflict and
a breakdown of
governance and security.

6.1. Hazard types

Ethiopia is subject to all the hazard types listed in the table above, though some are more frequent and
have more severe impacts than others. Figure 2 is taken from the priority country program document for
Ethiopia, under the Global Facility for Disaster Reduction and Recovery (GFDRR 2009:3). The graph on
the right shows that the most frequently occurring natural hazard in Ethiopia is flood, followed by
epidemic diseases and then drought. However, the table on the left shows that droughts continue to be
by far the most important type of disaster event in the country, in terms of the number of people
affected. ‘Drought’ in this context is primarily a short-hand term for food production and livelihood
failures (both in crop and livestock systems) caused by inadequate or untimely rainfall. Drought-induced
food crises are generally slow-onset disasters, although the speed at which their impacts are felt by the
affected population can vary widely, depending on production systems and vulnerability levels.

5 e.g. soil erosion, deforestation, and desertification.

10
TARGETING RELIEF FOOD ASSISTANCE PART TWO: CORE PRINCIPLES

Figure 2: Scale and frequency of disasters triggered by different hazards in Ethiopia

Drought therefore continues to be a major focus of Ethiopia’s disaster risk management system:
attention to other hazard types should not distract from this. At the same time, a forward-looking
emergency response system must be prepared for potential future hazards as well as familiar ones. The
frequency of extreme weather events (including floods) has increased in recent years, and it is expected
that climate change will also make the seasons increasingly unpredictable. Some hazards such as
drought, flood, human epidemics and livestock disease are considered likely to increase in both
frequency and severity. Exogenous market shocks (such as the global rise in food prices) can also trigger
acute food insecurity, especially when combined with other shocks or stressors.
Multi-hazard analysis supports a broader and more accurate understanding of the causes of, and
appropriate remedies for, food insecurity. Historically, the targeting of food aid at both geographical and
beneficiary levels was overly driven by a food balance-sheet model, which can be simplistically stated as:

consumption needs – (own production + stocks) = food availability deficit.

In reality, it has long been understood that rural communities in Ethiopia do not rely entirely on their own
crop production. They have complex and varied livelihoods, enabling them to access food in various ways
including purchase, exchange, collection of wild foods6 and gifts, in addition to consuming their own crop
and livestock produce.7 Any hazard affecting food production, livelihoods, social support, income
sources, market access or purchasing power (particularly for the poor and vulnerable who generally
produce less of their own food and are more market-dependent) may therefore have an impact on a
household’s food access. Therefore, change in any of the income and food sources of a community or
household – whether negative or positive– must also be factored into the equation, which can then be
restated as:

consumption needs – (own production + stocks + food purchased, gathered, received as gifts, etc.) = food access deficit.

6 Wild or gathered foods include plants that are uncultivated or common property, and animal products from hunting
and fishing. In some cases the use of wild foods is a distress strategy and/or potentially dangerous: in these
cases, it should not be considered part of normal food access. The acceptability of wild foods is specific to
locations and food types, and must be assessed in context.
7 WFP‟s definition of household food access is: “A household‟s ability to regularly acquire adequate amounts of food
through a combination of their own home production and stocks, purchases, barter, gifts, borrowing or food aid.”
(WFP (2005b). Emergency Food Security Assessment Handbook. Rome, World Food Programme)

11
TARGETING RELIEF FOOD ASSISTANCE PART TWO: CORE PRINCIPLES

The household food access deficit (or shortfall) is defined as “The difference between households’
nutritional requirements – what they need in order to re-establish or maintain satisfactory nutritional
health and to carry out productive activities – and what they are able to provide for themselves without
adopting distress strategies.” (WFP 2005b)

Since 2006 the Ethiopian early warning and needs assessment system has adopted an analytical
framework called the Household Economy Approach (HEA), which is able to model the impact of all
relevant hazards, singly or in combination, on the overall livelihoods and food access of households in
different areas and wealth categories across the country. This framework is central to the improved
targeting of food assistance. It is described further in Parts Four and Five.

6.2. Disaster types

Targeting is often easier (although logistics and other aspects may be much more difficult) in a fast-onset
disaster. This is partly because it is clear when the disaster event happened and therefore when the
response is needed. By contrast, in a typical slow-onset or cumulative crisis, conditions will gradually
worsen until a tipping-point or emergency threshold is reached. This tipping-point may be clear after it
has happened, but an effective relief response should begin before it is reached (that is, before lives and
livelihoods begin to be lost). This simple fact underlies the huge investment in early warning of food
crises in Ethiopia, and the long-established emphasis on (early, preventive) targeting by socio-economic
criteria in order to identify and support vulnerable people before they experience acute malnutrition,
destitution or starvation.8
Another reason why targeting is often easier in fast-onset disasters is that, depending on the type of
hazard that triggered the disaster, it is usually clear which people need assistance and which do not. For
example if a village is destroyed (whether by flood or conflict), everyone who has lost their homes and
possessions is likely to be in equal need of short-term food and non-food relief. If the same village suffers
a drought, or a disease outbreak that destroys crops or livestock, it is likely that everyone in the
community will be affected to some extent. However, some people will be able to cope with the impacts
and continue to feed their families by drawing on stocks and claims or by adapting their income
strategies. Not everyone will be in need of aid. Therefore, targeting criteria and systems are needed to
identify those who cannot cope without assistance.

6.3. Multi-sectoral response

Taking a multi-sectoral overview of response priorities provides a stronger basis for deciding when food
assistance is, or is not, an appropriate component of a disaster response. The first targeting question
after the nature of a disaster has been assessed should be: what kind of assistance do the people need?
Is short-term food access a problem? If so, what form of food assistance is most appropriate (in-kind or
cash/voucher), and what other kinds of assistance should accompany it?
In the past, food aid has sometimes been a ‘default’ response to disasters even when it was not the most
appropriate form of assistance in a given situation. This has been particularly common in pastoralist
areas, where other interventions (particularly actions to preserve or restore livestock resources) would
often be more effective in ensuring people’s food security. Accordingly the MoARD has issued detailed
guidelines for livestock relief interventions in pastoralist areas. The introduction notes that:
“Humanitarian assistance in pastoral areas has been dominated by food aid since
emergency interventions began in the 1970s, and food aid provision has been based on the
objective of saving human lives. However, it is increasingly recognized that emergency
assistance during drought or flood should also aim to protect people’s livelihoods. In
pastoral areas, livelihoods-based emergency programming means protection of pastoral

8 The complementary uses of socio-economic and nutritional targeting criteria are covered in Part Three, which
provides a technical toolbox of targeting methods.

12
TARGETING RELIEF FOOD ASSISTANCE PART TWO: CORE PRINCIPLES

livestock in appropriate numbers, and support to the services and markets which are
needed to assist rapid recovery.” (MoARD 2008:1)

 See MoARD 2008, National Guidelines for Livestock Relief Interventions in Pastoralist Areas of
Ethiopia

This does not mean that food assistance is not also needed in many emergencies, whether in pastoralist
or other livelihood zones. The important point is that food assistance should be targeted at the problems
it can solve, while other needs should be appropriately met by other sectors to ensure the recovery and
future resilience of disaster victims.
Box 1: Example of a disaster requiring a complex multi-sectoral response

In 2010, more than 86,500 people were temporarily displaced in parts of Afar, Amhara, Oromia,
SNNPR, Somali and Tigray in August and September due to flooding. In addition to emergency
food rations, the affected populations were found to be in need of shelter, non-food items, health
and nutrition, water, sanitation and hygiene, agriculture and livestock and education interventions.
Humanitarian Requirements Document July-December 2010 (DRMFSS 2010b:1)

In urban contexts, household food access is highly monetised and market-dependent. Therefore,
innovative modes of food assistance such as cash transfers (including employment provision through
cash-for-work), vouchers, or self-targeting price subsidies and controls should be considered. In-kind
food assistance may also be appropriate in urban food crises, depending on the cause and nature of relief
needs. The mode of assistance chosen will have implications for the appropriate targeting methods.
Because Ethiopia has little experience so far with urban relief operations, it is recommended that detailed
targeting guidance for urban emergencies should be added to the present guidelines in future as and
when lessons are learned.
Revisions or supplements to the operational parts of these guidelines may also be necessary as
experience grows with targeting food assistance in response to different kinds of disaster (such as floods
or market-induced food crises). However, the basic principles of targeting apply to all disaster and
response types. For non-food sectors, technical and targeting guidelines from the appropriate ministries
should be referred to as necessary.

7. Transitory-acute versus chronic food insecurity: defining the target

group for relief
Institutionally, the policy shift to DRM has been reflected in the establishment of the Disaster Risk
Management and Food Security Sector (DRMFSS) within the Ministry of Agriculture (MoA), following the
government’s wide-ranging business process re-engineering (BPR). The DRMFSS contains two parallel
directorates, as shown in Figure 3: the Early Warning and Response Directorate (EWRD), which replaces
the DPPA, and the Food Security Coordination Directorate (FSCD), which directs the Food Security
Programme (FSP). The FSP includes the Productive Safety Net Programme (PSNP), household asset-
building programme (HABP), complementary community infrastructure (CCI) programme and
resettlement. The most significant aspect of these changes for the present guidelines is the separation of
mandates for responding to chronic and transitory food insecurity, which dates from 2004 when the
PSNP was designed.

13
TARGETING RELIEF FOOD ASSISTANCE PART TWO: CORE PRINCIPLES

Figure 3: Separation of mandates for targeting transitory and chronic food insecurity

MoA

DRMFSS

EWRD FSCD
 Early Warning Food Security Program for
 Emergency Response chronic food insecurity
(including transitory /  PSNP  HABP
acute food insecurity)  CCI  resettlement

The PSNP, implemented in most regions since 2005, was specifically designed to remove from the relief
registers millions of chronically food insecure households, living in chronically food insecure weredas,
who had been receiving food aid year after year. This major innovation was in recognition that relief
food had been inappropriately used in the past to treat the symptoms of long-term structural poverty
and food insecurity, while doing little or nothing to address the causes. Instead, PSNP clients now receive
predictable seasonal payments in cash and/or food, in exchange for work on community development
projects (according to their labour capacity). This both frees and challenges the relief system to ensure
that its targeting is complementary, and is effectively focused on transitory (or transitory-acute) food
insecurity. The PSNP provides predictable seasonal transfers in pre-selected weredas for a stable
beneficiary group who are registered for several years, while the relief system provides short-term
emergency assistance for the acutely food insecure, whenever and wherever the need arises.
Some clarification is needed of the terms ‘chronic’ and ‘transitory’ food insecurity. Put simply, “chronic
food insecurity is a long-term or persistent inability to meet minimum food consumption requirements,
while transitory food insecurity is a short-term or temporary food deficit” (Devereux 2006:xi). Chronic
food insecurity is usually associated with underlying poverty, inadequate livelihoods and vulnerability,
while transitory food insecurity is usually caused by shocks or disaster events. Transitory food insecurity
has also been defined as “a sudden (and often precipitous) drop in the ability to purchase or grow
enough food to meet physiological requirements for good health and activity” (Barrett and Sahn 2001,
quoted in Devereux 2006:2), or “a crisis-induced reduction in food access” (WFP 2005b:83).
The PSNP defines its chronically food insecure client group in operational terms as “households who fail
to produce or purchase enough food even in times of normal rains” (PIM MoARD 2010p.7) - that is, they
cannot access sufficient food for their families, even in the absence of any disaster event such as an
agricultural drought. In contrast, “transitory food insecurity” is defined as “temporary needs that arise
from a specific and temporary shock” (ibid. p.30).
These terms (chronic and transitory) refer to the duration or time dimension of food insecurity.
However, as Devereux points out in his review of the complexities and confusions around this
terminology, “the intensity or severity of food insecurity is more important and urgent in an emergency
programming context than *its+ duration” (Devereux 2006:5, italics added). These guidelines therefore

14
TARGETING RELIEF FOOD ASSISTANCE PART TWO: CORE PRINCIPLES

use the term ‘transitory-acute’9 to define the target situations and groups for relief food assistance,
adopting WFP’s definition:
“Transitory-acute food insecurity ... [is] ...a situation of people and households who,
following a shock, are temporarily unable to meet their food intake needs without
sacrificing productive assets or undermining human capital.” (Emergency Food Security
Assessment Handbook, WFP 2005b:369).
Building on Devereux’s matrix combining the duration and severity dimensions of food insecurity, Table 4
offers descriptive definitions of moderate and severe transitory food insecurity. The target group for
relief food assistance can be conceptually defined as people in the transitory + severe (or ‘transitory-
acute’) situation, outlined by the bold red box on the table. The definition given in this box clarifies the
important point that relief assistance should not be provided to everyone who has been affected by a
disaster: targeting must consider their ability to cope.

Table 4: ‘Composite food insecurity’ matrix: combining duration and severity

SEVERITY 

Moderate Severe

Chronic (moderate chronic food insecurity) (severe chronic food insecurity)

DURATION 

Transitory People‟s access to food (through People‟s access to food (through purchase,
purchase, production or other means) production or other means) is temporarily
is temporarily reduced by a shock or a reduced by a shock or a slow-onset disaster,
slow-onset disaster,
and they are unable to maintain an adequate
but they are able to maintain an nutritional intake,
adequate nutritional intake through
or they are able to maintain an adequate
sustainable coping strategies and/or
nutritional intake only by resorting to
support from ongoing development or
unacceptable or damaging coping strategies,
social protection programmes, without
risking irreversible damage to their health or
irreversible damage to their health or
livelihoods.
livelihoods.

Emergency food assistance needed

Matrix adapted from Devereux 2006:7

For targeting purposes it is important to understand that ‘chronic food insecurity’ and ‘transitory-acute
food insecurity’ are two different situations: they are not two separate categories of people. People who
are chronically food insecure nearly always have limited resilience or coping capacity and therefore are
most likely to suffer acute food insecurity when a disaster strikes. Thus, chronic food insecurity is often
an underlying contributory cause of acute food crises, particularly in slow-onset disasters.
Repeated periods of transitory food insecurity can also erode assets and can cause, or worsen, chronic
food insecurity. Figure 4 illustrates how the same series of shocks can have different impacts on
households’ livelihoods, depending on their asset levels and overall resilience. The vulnerable
households (on the left) are at risk of being pushed into chronic food insecurity by depleting their assets
in order to survive repeated shocks. The resilient households (on the right) may suffer moderate
transitory food insecurity due to the same shocks, and may draw down their resources in order to cope
with the impacts, but they are able to recover. In principle, it is the vulnerable households who should be

9 The basic meaning of transitory is „passing‟ or „temporary‟. The basic meaning of acute is „severe but of short
duration‟.

15
TARGETING RELIEF FOOD ASSISTANCE PART TWO: CORE PRINCIPLES

targeted for relief food assistance in an emergency response, not only to relieve their suffering but also
to prevent them becoming chronically food insecure or destitute (and therefore requiring increasing
quantities of aid in the future).

Figure 4: Effects of shocks on the resources of vulnerable and resilient households

Source: Devereux 2006:9

The relationship between chronic and transitory food insecurity is recognised by the PSNP’s Programme
Implementation Manual (PIM). It states that transitory or temporary needs, caused by an increased
deficit in food access following a shock, can affect either chronically food insecure households or those
who are normally food secure. In a disaster, a PSNP client household may already be adequately
supported by their regular safety net transfers. If not, they may need either additional rations (if all
family members were not previously registered) or additional months of support (beyond the 3, 6 or 9
months’ transfers which the PSNP aims to provide during the normal hungry season).
In its current phase, the PSNP is designed as a scalable safety net, meaning that its size can be adjusted in
response to changing needs. When there is an unexpected, temporary rise in food needs within PSNP
weredas, the programme will first access the Contingency Fund and then, if this is not sufficient, it will
activate the Risk Financing Mechanism to provide additional short-term food assistance. The target
groups for contingency and Risk Financing support are households within the programme weredas who
cannot meet their basic food needs in the short term, whether or not they are PSNP clients. If and when
the additional (transitory-acute) needs exceed the capacity of the Risk Financing Mechanism, then an
emergency relief response is needed. 10
In weredas where the PSNP does not operate, there is currently no risk financing mechanism in place: the
relief system therefore needs to act immediately when acute food needs have been identified due to a
fast-onset or slow-onset disaster. Slow-onset disasters, as mentioned above, are often made worse by
pre-existing chronic food insecurity. Fast-onset disasters, on the other hand, can cause acute temporary
food insecurity, whether or not the victims were previously food secure. In either case, it is the present
situation of the people affected that determines whether or not they need relief food assistance.
To sum up, relief food assistance is an appropriate response to food insecurity that is both transitory and
severe (or ‘transitory-acute’): that is,
 transitory - the need for assistance is expected to be of finite duration (not permanent or
cyclical), and

10 Full operational rules for the PSNP can be found in the PIM (MoARD 2010) and Risk Financing Guidelines
(MoARD 2009).

16
TARGETING RELIEF FOOD ASSISTANCE PART TWO: CORE PRINCIPLES

 severe - without assistance, people will suffer significant or unacceptable negative impacts (on
health, child growth and development, and/or livelihood viability). In severe situations, acute
food insecurity can lead to starvation and ultimately famine. A well-targeted responsive relief
system should be committed to saving lives by intervening early in a developing crisis, before it
reaches such a level.
Chronic food insecurity requires longer-term remedies. However, because chronic food insecurity can
underlie or deteriorate into acute food crises, no-one should be excluded from relief on the grounds that
they belong in a ‘chronic’ category. If acute food needs are found to have long-term underlying causes
which are not being addressed, people’s immediate needs should first be met with short-term relief
assistance, and then development or safety-net actions should be put in place to tackle the roots of the
problem.
It follows from the definitions in this section that eligibility for relief should depend on the severity of the
disaster impact and the resulting size of a household’s food access deficit, balanced against the
household’s coping capacity (that is, their ability to fill the deficit without additional assistance and
without unacceptable negative consequences). Practical details of how these principles should be
applied, and how relief targeting should be harmonised with PSNP targeting at the household level, are
given in Part Five of these guidelines.

Box 2: International Standards for Targeting in Disaster Response

a) The Sphere Project

Key indicators
1. Targeting criteria must be based on a thorough analysis of vulnerability.
2. Targeting mechanisms are agreed among the affected population (including
representatives of vulnerable groups) and other appropriate actors. Targeting criteria are
clearly defined and widely disseminated.
3. Targeting mechanisms and criteria should not undermine the dignity and security of
individuals, or increase their vulnerability to exploitation.
4. Distribution systems are monitored to ensure that targeting criteria are respected and that
timely corrective action is taken when necessary.
Sphere Project 2004:35-7
Humanitarian Charter and Minimum Standards in Disaster Response
http://www.sphereproject.org

b) The United Nations (World Food Programme)

“Assistance will be guided solely by need and will not discriminate in terms of ethnic origin,
nationality, political opinion, gender, race or religion. ...[A]ssistance will be targeted to those
most at risk from the consequences of food shortages, following a sound assessment that
considers the different needs and vulnerabilities of women, men and children.”
WFP Humanitarian Principles (WFP/EB.A/2004/5‐C), quoted in WFP 2006:7

17
TARGETING RELIEF FOOD ASSISTANCE PART TWO: CORE PRINCIPLES

8. International minimum standards for relief targeting

The DRM Policy states that “international humanitarian principles and minimum standards of operations
will be taken into account and adapted as per the local conditions.” Accordingly, the relevant
international standards for targeting are set out in Box 2. Firstly, minimum standards for targeting in
general (across all technical sectors in humanitarian operations) are quoted from the Sphere Project
Handbook. These are the most widely used and thoroughly tested international standards for disaster
response, developed since 1997 by an international group of humanitarian agencies. Ethiopia has already
adopted the Sphere standards for WASH (Water, Sanitation and Hygiene) interventions in emergencies.
The second extract in Box 2 is WFP’s statement of fundamental targeting principles, specifically in
relation to food assistance.
Both formulations stress the core principles that priority should be given to those most at risk, and that
targeting must be needs-based (following an impartial and transparent assessment of the vulnerability
and needs of different groups). All actors and organisations involved in the targeting of relief food
assistance in Ethiopia should strive to meet these standards.

9. Guiding principles of the DRM Policy applied to targeting

The DRM Policy (Section 2.3) sets out thirteen guiding principles which should characterise the whole
cycle of disaster risk management, including emergency response. All of these have direct relevance to
the targeting of relief food assistance, as summarised in the table below. The principles are listed here
with the same order and numbering as in the policy document, for ease of cross-reference: this is not
necessarily the order of importance for targeting.

Table 5: Guiding Principles of the DRM Policy Applied to Targeting

Principle Application to the targeting of relief food assistance

1 Integration into Proper targeting of relief protects development gains at household and
development national level by mitigating the negative impacts of food crises, i.e.
impoverishment, increased chronic food insecurity, loss of livelihoods
and damage to „human capital‟ (through impaired health, growth or
education).

2 Protection of natural Timely and adequate relief targeting should enable vulnerable
resources households to avoid using environmentally damaging coping strategies.

3 Decentralized and Relief targeting decisions are increasingly decentralized to regions, and
community-based 11 will be increasingly tailored to local calendars and needs as the regular
wereda-based early warning system is strengthened in the coming
years.
Decentralized prepositioning of stocks and wereda DRR contingency
planning will improve the timeliness of relief targeting.
Within disaster-affected weredas, communities have a key role in
deciding the final level of targeting (registering households and
individuals to receive aid). These guidelines aim to strengthen the
empowerment of communities through more consultation, better
information flows, and respecting community values in tailoring targeting

11 “DRM systems shall be decentralized and community-based whereby communities play decisive roles in the
planning, execution, monitoring and evaluation of disaster risk management projects and programmes.”(FDRE
2010a, DRM Policy 2.3.3.)

18
TARGETING RELIEF FOOD ASSISTANCE PART TWO: CORE PRINCIPLES

criteria and processes to the local context.

4 A participatory Successful targeting outcomes require a chain of actions by many

approach 12 different institutional actors throughout the disaster-response cycle.
Therefore, participation by various levels and sectors of the government,
humanitarian partners, communities and other stakeholders in the
targeting process is highly valued and should be strengthened.

5 Accountability and All actors involved in targeting relief assistance have an ultimate
responsibility responsibility to the potential beneficiaries and disaster victims as well
as to the authorities above them.
Detailed accountability chains for targeting in disaster response should
be set out in the DRM Policy legislation which will be developed after the
approval of the Policy. Each Region can also set up its own measures
to effect accountability according to their specific situations and
institutional structures. The lines of accountability should include any
humanitarian partners who are operationally involved in relief targeting,
and should be adjusted as necessary for specific emergency operations.
Accountability requires systematic record-keeping and reporting of
targeting decisions and actions.

6 Protection of human The updated policy restates the bottom-line commitment of the 1993
life NPDPM: “No human life shall be lost for want of basic relief assistance
in times of disaster.”
Relief targeting should therefore be adequate and timely, providing
sufficient support as early as possible in a disaster situation to prevent
or minimise any loss of life.

7 Information is power Information on targeting criteria, procedures and outcomes should be

recorded and freely available, to ensure accountability and learning.
Analytical reports on the needs assessment findings (which provide the
planning figures for geographical and household targeting) should be
circulated, to improve transparency and strengthen the logical links
between geographical and household targeting.
Communities should be empowered by the provision of information
about the quantities and timing of relief deliveries, the needs
assessment findings for their area, and the resulting provisional
targeting criteria.

8 Ensuring basic needs Adequate13 food is one of the most basic human needs and rights.
Proper targeting of relief food assistance aims to ensure that no citizen
of Ethiopia will face severe hunger or starvation, or lose their ability to
meet their basic needs in the longer term, because of disasters.

9 Impartiality Targeting should be based only on need, after an unbiased impartial

assessment considering the impacts of the disaster and the different
vulnerabilities and coping capacity of different groups in the disaster-
affected population.

12 “DRM Systems will ensure the proper participation of all stakeholders including the private sector, academic and
research institutions, humanitarian organizations, etc.” (FDRE 2010a, DRM Policy 2.3.4.)
13 „Adequate‟ refers to both the quantity and nutritional quality of food. Food security requires “sufficient safe and
nutritious food for a healthy and active life” (see Glossary).

19
TARGETING RELIEF FOOD ASSISTANCE PART TWO: CORE PRINCIPLES

10 Respect cultural Criteria and procedures for household targeting should be adapted to
norms the cultural norms and values of the disaster-affected community,
through a process of consultation. However, if cultural norms clash with
the principles of needs-based targeting, the principles take precedence.

11 Minimize dependency Relief targeting should be limited to the areas, people, and periods
and promote livelihood where it is needed, in order to avoid dependency and disincentive
recovery effects as well as the waste of scarce resources.
Relief targeting should be adequate to support beneficiaries through the
period of acute need without irreversible loss of assets, in order to
enable early livelihood recovery and prevent more people becoming
chronically food insecure and therefore dependent on aid in the future.

12 Give due attention to Targeting of relief food assistance should always prioritise the most
the most vulnerable vulnerable, and must consider:
groups
economic vulnerability (people who cannot cope with the disaster from
their own resources or income);
social vulnerability (people who may be marginalised from informal
social protection systems or have limited support from their community
or family, due to gender, age, ethnicity or other reasons); and
nutritional vulnerability (people with increased or special nutritional
needs, such as children, pregnant and lactating women, the elderly and
people living with HIV/AIDS).

13 Avoid malpractices Deliberate mis-targeting or diversion of relief resources should be

subject to legal sanctions, to be detailed under the planned review of
legislation relating to the DRM Policy.14
Robust systems for monitoring and reporting must be in place to detect
and correct any targeting malpractice.

In later sections of the guidelines, linkages to the core principles are highlighted by this link
symbol.

14 The DRM policy states “There shall be a review and, where appropriate, revision of existing legislation ...
Responsible individuals and bodies shall be empowered and capacitated to implement risk management activities
and be held legally accountable in case of failures of discharging their DRM duties and responsibilities.”

20
TARGETING RELIEF FOOD ASSISTANCE PART THREE: TOOLBOX

PART THREE: THE TARGETING TOOLBOX

10. Six key questions of targeting

The purpose of this ‘toolbox’ chapter is to provide users with the necessary technical concepts and
terminology for a clear and common understanding of issues that arise in the implementation of
targeting. ‘Targeting’ itself has a range of definitions which emphasise different aspects. For example,
“restricting the coverage of an intervention to those who are perceived to be most at risk, in order to
maximise the benefit ... whilst minimising the cost” (Jaspars and Young 1995:136), or “directing a
particular type or quantity of food, to a defined population group” (Taylor and Seaman 2004:4). Since
these guidelines are primarily concerned with how this can best be done, they use a practical, operational
definition:
“A targeting system comprises mechanisms to define target groups, to identify members of the
target groups and to ensure that assistance reaches intended beneficiaries and meets their needs”
(WFP Programme Design Manual, italics added).
Defining target groups is primarily the task of early warning and needs assessment, while reaching them
depends to a large extent on distribution logistics and management. Detailed technical guidelines exist
for these activities, and therefore this document focuses mainly on the process of identifying the people
who are eligible for relief food assistance (and subsequently verifying whether they received it).
However, a targeting system cannot be considered successful unless all three components are effective
and inter-linked with each other.
The overall targeting system (including vulnerability and needs assessments) needs to answer six key
questions, which together provide a logical framework for targeting decisions:
1. WHO needs relief? (that is, the definition and general characteristics of target groups)
2. WHY? (analysis of the disaster event and its impact, combined with vulnerability assessment)
3. WHERE? (the location and concentration of those in need, for geographical targeting)
4. WHEN and for HOW LONG? (the seasonality of relief needs, and the expected duration of the
problem)
5. WHAT and HOW MUCH? (the type and quantity of relief needed – including the appropriate balance
of food and non-food assistance)
6. HOW can the people in need best be identified and reached? This question involves a choice of:
o Distribution modality/ies (e.g. general food distribution; cash; vouchers; supplementary feeding;
school feeding; C/FFW);
o Level of targeting (e.g. geographical targeting only, or selection of groups, households or
individuals);
o Criteria or indicators which characterise the target group members and can be used to identify
them (e.g. nutritional, economic, or demographic criteria; direct measurement or proxy
indicators); and
o Targeting method (e.g. self-targeting, administrative, community-based, or institutional
targeting).
The following sections focus on the last three points of the “how” question, looking in more detail at
targeting levels, criteria and methods. These sections provide a technical overview of the range of
options that are available when designing and implementing a targeted relief operation. Instructions on
the methods and procedures to be adopted under current Ethiopian policy are described in Parts 4 and 5.

21
TARGETING RELIEF FOOD ASSISTANCE PART THREE: TOOLBOX

11. Levels of targeting

Target groups can be defined and identified at different levels of aggregation or detail - at geographical,
group, household or individual level.
Geographical targeting identifies areas where a disaster event has occurred and assistance is needed. It
can be applied at various scales, from very large areas (for example, when international donors select
countries to focus their resources on) to very small ones (selecting kebeles, villages or camps). Areas for
geographical targeting may be defined by administrative boundaries, or by analytical units such as agro-
ecological or livelihood zones.
In fast-onset or localised emergencies, geographical targeting (i.e. identifying which areas need assistance
and which do not) may be relatively straightforward. If there is an earthquake or a flood, it is easy to see
the boundaries of the damage. In widespread slow-onset emergencies (such as drought-induced food
crises), geographical targeting can be less clear-cut as there is often a continuum of need, with different
areas affected to different degrees. The main disadvantage of geographical targeting is that it
automatically excludes food-insecure or vulnerable people who happen to live outside the targeted
areas. Because of this, it can produce inbuilt exclusion errors. In some circumstances - for example, in
emergency situations when relief is provided to one wereda or kebele but not to the neighbouring one -
the exclusion of needy people who are outside the targeted area can be a contentious issue.
In some relief operations, geographical targeting is the only level of targeting used. Once an area of
manageable size has been targeted (for example, a district, livelihood zone or community), aid is
provided to everyone in that area on a blanket basis, without any further screening. Examples of
situations where this may be appropriate are:
 Acute emergency situations where immediate action is needed, and screening is not feasible or
cannot be done fast enough. In such cases a blanket distribution may be provided for a short
period, followed by more detailed targeting if the emergency continues.
 Types of disaster where it is clear that the whole area or community is in need (e.g. fast-onset
disasters that have destroyed a whole community’s homes or livelihoods; initial stages of
displacement when everyone has lost access to their normal food and income sources).
 Insecure or inaccessible areas, where detailed targeting may be too dangerous for implementing
agencies or may exacerbate conflict. This is particularly relevant in complex emergencies.
 Some models of community-managed targeting (see below), in which it is agreed beforehand
that everyone within the targeted community will receive assistance.
In most cases, however, geographical targeting is the first stage in a process and is followed by more
detailed stages of targeting to prioritise the neediest groups, households or individuals within the
targeted areas.
Group targeting identifies sections of the population, based on an objective analysis that these groups
are particularly vulnerable or food-insecure in the given situation. Individual assessment of group
members is not needed for this level of targeting – usually the groups are defined by observable
characteristics (such as gender, age, or disability) or agreed social categories (such as widows and
orphans).
Household targeting takes the household as the smallest social and economic unit, and provides
resources to be shared among household members. Eligible households may be identified by:
 the overall household situation (e.g. total household income, assets or food access);
 the status of the household head (e.g. age, gender, health, occupation); or
 the presence in the household of a particular target group member (e.g. a malnourished child or a
chronically ill adult).
Since the definition of a ‘household’ varies in different social and economic contexts, it is important to
establish a shared understanding of the term with the beneficiary community before implementing

22
TARGETING RELIEF FOOD ASSISTANCE PART THREE: TOOLBOX

household targeting. For relief food assistance, the international standard practice is to define the
household as a woman and the people she cooks for. This can be applied in most socio-cultural contexts,
and has a number of advantages (see Section 22).
Individual targeting, in food security contexts, is most often applied in nutritional and health
programmes that screen beneficiaries through direct measurement (such as weight-for-height and other
anthropometric indicators) or medical referral (see section 12.1.a.).

12. Choosing targeting criteria

Targeting (or ‘eligibility’) criteria are the characteristics or measurements by which target group members
are identified. The choice of targeting criteria can significantly affect the success in reaching the intended
target groups, and the feasibility of implementing the targeting process. Box 3 summarises some
desirable qualities of targeting criteria.

Box 3: Characteristics of good targeting criteria

In general, targeting criteria should be:

 logically linked to vulnerability and needs analysis
 sensitive (that is, they will correctly include target group members);
 specific (they will correctly exclude non-target group members);
 feasible (given the time, resources and capacities available);
 acceptable to the beneficiaries and their community; and
 verifiable, so that targeting decisions can be checked and challenged by monitoring staff or
community members.

Targeting criteria may be either direct measurements of target group characteristics (for example,
anthropometric measures of nutritional status, or individual means-testing of total asset and income
levels), or they may be ‘proxy’ (indirect or substitute) indicators which are taken to represent more
complex characteristics. For example, housing quality, ownership of specific assets, or the gender of the
household head may be used in some situations as proxy indicators to represent poverty or food
insecurity. Proxy indicators are generally easier and cheaper to apply than direct individual
measurements, but good data and analysis are needed to ensure that they are reliably associated with
the target group definition and will therefore accurately identify the right people, without large inclusion
or exclusion errors.

12.1. Linking targeting criteria with vulnerability and needs analysis

As outlined in Section 10, targeting should be linked to vulnerability analysis, risk profiling and livelihood
baselines which provide a diagnosis of why certain people are vulnerable, and what they are vulnerable
to. When a disaster strikes, needs assessments then identify who has been affected and needs
assistance; what kind of assistance is needed, where, when and how much. All these pieces of
information are essential to compile a clear definition of target groups. Clear definitions in turn help to
generate appropriate targeting criteria and methods which will ensure that those people receive
assistance. "The question of targeting goes hand in hand with assessment, as it raises the same
conceptual questions about needs and vulnerabilities, and also very practical issues of how to identify
and reach particular groups. Identifying target groups raises the fundamental question 'who is at risk
and why?', or alternatively 'who is vulnerable to what?'" (Jaspars and Young 1995:33).

23
TARGETING RELIEF FOOD ASSISTANCE PART THREE: TOOLBOX

SPHERE Key Indicator 1:

“Targeting criteria must be based on a thorough analysis of vulnerability.”
(see Box 3 in Part Two)

Clarifying the types of vulnerability which are causing acute food insecurity in a given situation will
therefore help to determine the types of criteria which will be most effective in identifying and reaching
the people in need. Some examples are given in Table 6. The following sections then describe the most
common sets of criteria used in the targeting of food assistance, i.e. socio-economic, nutritional, and
categorical criteria.

Table 6: Types of vulnerability and appropriate targeting criteria

Type of vulnerability Appropriate targeting criteria (examples – not a complete list)

ECONOMIC vulnerability  crop and livestock conditions

 employment opportunities
 assets, stocks and income
 sources of support (e.g. remittances)

PHYSIOLOGICAL vulnerability  age (e.g. children under 5; elderly)

 pregnancy and lactation
 illness (including HIV/AIDS, malaria, TB and other epidemic diseases)

SOCIAL vulnerability  people with limited family support, e.g. widows, orphans
 socially excluded or marginalised groups (locally defined)

POLITICAL vulnerability  refugees or displaced people (criterion: place of origin)

 communities and individuals exposed to violence or marginalisation (locally
defined)

12.1.a. Nutritional criteria for prevention or cure of malnutrition

Measurements of nutritional status are among the most objective and standardised indicators for
targeting individuals who are currently malnourished, or (at geographical level) areas with high levels of
malnutrition. Technical guidance on the collection and analysis of nutritional (anthropometric) data and
the threshold values to be used for targeting can be found in the Emergency Nutrition Co-ordination
Unit’s guidelines:

 See ENCU 2004, Emergency Nutrition Intervention Guideline (Updated version due 2011)

Nutritional criteria are sometimes used as proxy indicators for household food insecurity. However, this
can be quite inaccurate as malnutrition can have complex causes apart from shortfalls in food access or
availability. Malnutrition is also a late (outcome) indicator of food insecurity, and therefore targeting by
nutritional criteria alone may cause a late response which leaves livelihoods unprotected.

24
TARGETING RELIEF FOOD ASSISTANCE PART THREE: TOOLBOX

12.1.b. Economic criteria for early response and livelihood protection

The use of economic criteria (such as income level, assets, wealth-ranking and livelihood profiles) to
target relief food assistance is based on the analysis that food insecurity is fundamentally related to
poverty. People are vulnerable to acute food insecurity, under-nutrition and at worst starvation because
their sources of food and income are limited and unreliable, and their stocks of assets or claims on other
people do not give them enough to fall back on in a crisis.
Therefore, an early response, especially to slow-onset disasters, is usually most effectively targeted by
economic criteria. Identifying and assisting people who are economically vulnerable and at risk of
eroding their livelihoods in order to meet short-term food needs can prevent a food crisis developing into
a potential famine situation, reduce the risk of severe malnutrition, and protect livelihoods for the future.
Unfortunately economic criteria are the most difficult to apply by objective measurement, and often raise
the most disagreements when applied by other means (such as community targeting). In diverse
economies like Ethiopia’s they cannot easily be standardised, because incomes are impractical to
measure directly and proxies such as livelihood profiles and asset levels vary greatly between livelihood
zones. The HEA baselines have helped with this problem by providing reference values for asset levels
and sources of food and income by wealth group and livelihood zone, which may be used as provisional
or suggested targeting criteria.

12.1.c. Social and categorical criteria for targeting vulnerable groups

Nutritional and economic criteria are generally ‘continuous variables’ with a range of possible values,
meaning that a threshold or cut-off value must be decided in order to use them for targeting. For
example, if productive assets such as land or livestock are used as targeting criteria, a decision must be
made about how much land, or how many livestock, will qualify or disqualify someone. With nutritional
measurements the thresholds are more easily standardised (e.g. 12 cm MUAC for TSF), but they still
require measurement and there can be uncertainties or disputes about people who fall just inside or
outside the threshold.
Categorical (yes-no) indicators are an alternative or complementary set of targeting criteria which do not
need thresholds or measurement. Generally they are ‘observable characteristics’ which are easily
verifiable and easily agreed by community members. Examples are gender, age,15 pregnancy, disability,
and social or life-cycle situations such as being widowed or orphaned.
Certain demographic groups (children under five, women who are pregnant or breastfeeding, and the
elderly) are physiologically more vulnerable to malnutrition than others, and are therefore often targeted
with supplementary foods without nutritional screening.
Categorical indicators can be easier to apply than economic and anthropometric criteria, but they should
not be used simply for that reason. The choice of targeting criteria should always be based on evidence
and logic. Certain groups (women, children and the elderly) are more vulnerable in most emergency
situations and should therefore be given special attention in targeting, to ensure their needs are met. If
other categories or groups of people have been assessed as particularly vulnerable in a given disaster,
then those categories may also be used as targeting criteria.

DRM Principle 12: Give due attention to the most vulnerable groups
(see Table 3 in Part Two)

15 Technically, of course, it is a simplification to regard age as a „category‟, but in practice communities find it
relatively easy to agree about who is „elderly‟.

25
TARGETING RELIEF FOOD ASSISTANCE PART THREE: TOOLBOX

12.2. Combining targeting criteria

Except for nutritional targeting, and unless incomes can be directly measured, it is generally advisable to
use more than one targeting indicator to ensure that all the neediest are included and most of the people
who can cope without relief are excluded. This is because each single proxy indicator is likely to
introduce a margin of error in identifying the target group. If appropriate indicators are combined, these
errors can cancel each other out to a large extent. Single indicators, especially those based on assets,
also run the risk of creating dependency traps (see Section 21.3).
One of the simplest approaches to combining indicators, which is particularly suitable for use in
community targeting (see Section 13.2.), is to agree a set of inclusion criteria on the one hand, and
exclusion criteria on the other. Then the intersection or convergence of the different indicators is used to
decide whether a particular household is eligible or not. For example, inclusion criteria may include
categories such as female-headed households, economic thresholds such as a locally-defined level of land
or livestock, and nutritional indicators such as a malnourished child in the household. Exclusion criteria
could include alternative income sources such as remittances or adequate support from development or
safety-net programmes.

13. Targeting methods and mechanisms

13.1. Administrative targeting

In administrative targeting (sometimes called ‘indicator-based targeting’), beneficiaries are selected:
 by outsiders such as government employees or humanitarian agency staff, acting in a professional or
official capacity, who are not themselves potential beneficiaries;
 using pre-defined criteria that should be objective, measurable and standardized.
Administrative or indicator targeting can be adapted to any level of targeting. For example, geographical
targeting of disaster-affected and food insecure areas is done administratively, based on analysis of early-
warning indicators such as production forecasts, rainfall, market trends, or prevalence of malnutrition.
At the household or individual level, administrative targeting can be applied to a wide range of criteria. It
may use direct individual measurement of target group characteristics, or it may employ ‘proxy’
indicators (see above). If proxy indicators are used, their validity and the appropriate threshold values to
determine eligibility for assistance should be based on a sound analysis of local data. Therefore, effective
administrative targeting requires a high level of capacity for data collection and analysis, as well as robust
and accountable systems for the actual implementation and recording of screening procedures.

13.2. Community targeting

Community (or community-based, or participatory) targeting is only applicable within the community, at
group, household or individual level. In this method, beneficiaries are selected:
 by ‘insiders’, meaning members of the community who are potential or actual beneficiaries;
 using a flexible combination of agreed criteria (defined or fine-tuned by community discussion) and
judgement of the overall relative need of different community members.
This approach is widely used in development and relief programmes, particularly where resources or
capacity are not adequate for reliable administrative targeting, or where community empowerment is an
objective in itself. Selection is based on the decision-makers’ prior knowledge and understanding of their
neighbours’ situation: thus, quantitative data collection and analysis by outsiders are not needed.
Participatory methods such as wealth-ranking are often used in community targeting decisions:
alternatively, the community may simply be asked to identify the poorest or neediest households. The
combinations of criteria used by community members tend to be more subjective and more complex
than in administrative targeting, although they usually include similar factors(such as assets, livelihoods,

26
TARGETING RELIEF FOOD ASSISTANCE PART THREE: TOOLBOX

age, gender, food stocks, or household size). Community-defined criteria and thresholds may also be
locally specific, and therefore more accurate in identifying the needy but also more difficult to
standardise or compare across areas.
Procedures and institutional forms for community-based targeting vary widely, depending on local
culture, social structures and decision-making institutions. Successful community targeting may be
carried out by established leaders, specially elected committees or representatives, or by the whole
community in an open meeting, again depending on local norms and power structures.
The term ‘community targeting’ can also cover a wide spectrum of degrees of community autonomy,
from complete decision-making power over how assistance should be distributed (with or without
targeting), to simply implementing pre-determined instructions.
International experience suggests that the conditions necessary for successful community targeting
include a stable and peaceful situation without extreme stress or serious inter-community conflict, and
well-established representative community structures for governance and decision-making. In some
emergency situations involving conflict or displacement, or in situations when it is desirable to target
vulnerable groups who are marginalised from the community, community targeting may not be
appropriate.
In addition to these contextual factors, four characteristics of the process are necessary for successful
community targeting:
 transparency – everyone in the community should know how targeting decisions were made;
 information - community members should know how much assistance has been allocated to their
area, and the rules for targeting it;
 accountability – there should be a channel for decisions to be challenged, and for representatives to
be replaced; and
 audit - by an impartial outside authority (whether from government or a partner organisation), which
can arbitrate disputes.

13.3. Self-targeting
The basic mechanism of self-selection (or ‘self-targeting’) is that beneficiaries themselves decide whether
or not to participate in an assistance programme, depending on the balance between the benefits they
can receive and the costs they are required to pay. Designing a programme so that the intended target-
group will self-select and others will choose not to participate therefore “requires choosing a benefit
which only the target population wants, or including a cost which only the target population is willing to
pay” (Lundberg and Diskin 1994:5).
 The benefit side of this balance is basically the value of the assistance offered (in cash, exchange or
nutritional terms). This value may be adjusted by offering a low-priced foodstuff that the better-off
will not want,16 or (for FFW/CFW) setting a payment rate that is below the market wage for
alternative employment.
 Cost factors for beneficiaries may include the time, effort and transport expenses needed to collect
rations; work requirements in the case of FFW / CFW; or the social costs of identifying oneself as
needy.
In Ethiopia, the use of self-targeting in employment-based programmes (food-for-work, cash-for-work,
employment generation schemes, or employment-based safety nets) has been thoroughly tested over
several decades. In theory, wages for such programmes should be set below the prevailing market rate
for unskilled labour so that only the poor will self-select, while people who can earn more in alternative
employment or who place a higher value on leisure time will choose not to participate. However, this

16 Commodities provided as relief assistance should, of course, always meet international quality standards and be
acceptable to the beneficiaries.

27
TARGETING RELIEF FOOD ASSISTANCE PART THREE: TOOLBOX

mechanism only works effectively if alternative employment is actually available in sufficient volume at
the time and place of need, and if better-off households can fully employ their labour in such better-paid
work. If these conditions are not met, there is little incentive for better-off people to self-exclude from
the programme. A review of the evidence during the design of the PSNP therefore concluded that “self-
targeting based on the wage-rate is not applicable since alternative employment opportunities are either
non-existent or minimal, especially in chronically food-insecure areas of the country” (2004 version of
PSNP PIM, section 4.2.1). This is equally or more true in emergency situations.
Other limitations of pure self-targeting in employment programmes are:
 Market wages in food-insecure areas, particularly in times of crisis, may be so low that a below-
market payment rate would not meet beneficiaries’ basic food needs. This would undermine the
primary objective of providing adequate food access to save lives and livelihoods, and would be
unacceptable on humanitarian grounds.
 Employment-based programmes automatically exclude or disadvantage individuals who cannot work
(such as the ill, elderly, pregnant, disabled or malnourished) and labour-poor households.
Alternative channels of support must be provided for these vulnerable groups.
Because of these limitations, and although the inherently self-targeting elements of work programmes do
encourage a minority of better-off households to self-exclude (see Sharp et al. 2006:37), self-targeting is
usually combined with administrative or community targeting systems to ensure that benefits reach the
intended target groups. Pure self-targeting which successfully reaches the intended target group without
large inclusion or exclusion errors can be very difficult to achieve, particularly in emergency situations
when the demand for assistance is likely to exceed the supply.
A different kind of self-targeting, which relies more on social factors than on cost-benefit judgements,
happens in successful community-based targeting. When targeting decisions are genuinely democratic
and consensual, they require the relatively better-off members of the community voluntarily to self-
exclude by agreeing that others are in greater need. Conversely, those in need of assistance should feel
empowered to self-target by putting their case forward to the community (or to appeal if they think they
have been unfairly excluded). In some countries, this kind of self-targeting is managed in a more
administrative mode: when a relief operation is announced, people who need assistance and believe they
are eligible are required to apply to the relevant local authorities, who then assess their claim against
predetermined criteria before deciding whether or not they should be registered. For this type of system
to be effective in reaching the intended target group, there must be universal access to information
about the assistance available and the eligibility criteria (for example, via radio broadcasts), as well as
robust, accountable administrative systems.

13.4. Institutional targeting

Institutional targeting methods provide food assistance to groups or categories of people who are
eligible, or already registered, for other services such as health or education. The most relevant
examples in acute food emergencies are school feeding, mother-and-child health (MCH) programmes,
and clinics providing medical treatment for patients such as those living with HIV/AIDS (PLWHA). The
targeting effectiveness of these approaches depends on the geographical coverage, and the
inclusiveness, of the existing institution. These factors are very variable and should be assessed in the
local context. For example, it has been found in some pastoralist societies that children from poor and
food-insecure households are more likely to attend school during food crises, and can therefore be
reached through expanded school feeding.17 However, in most cases children enrolled in school are
more likely to be from better-off families: therefore, school feeding would not reach the neediest
households (WFP 2005c).

17 In pastoralist areas of Northern Kenya, for example, rising school attendance figures are used as an early warning
indicator.

28
TARGETING RELIEF FOOD ASSISTANCE PART THREE: TOOLBOX

Table 7: Advantages and disadvantages of different targeting methods

ADVANTAGES DISADVANTAGES

+ Objective, standardised and verifiable (when – High costs in management time, data collection and
successful) analysis, screening procedures and monitoring.
ADMINISTRATIVE

+ Can be very accurate, therefore effective at – May not be feasible where administrative and
minimising errors (especially inclusion errors) information-processing capacity is low.
+ Appropriate for targeting by measurable criteria, – Risk of bias, corruption, intimidation, theft or error –
such as anthropometry for targeted supplementary requires monitoring / auditing.
feeding

+ Community members already know each other‟s – Community ideas of equity and vulnerability may
situation (assets, income sources, household size not match government and donor targeting priorities
etc.), so no need for costly data collection and – Community may disagree with the principle of
analysis targeting, and prefer to share aid among everyone
+ Promotes participation and ownership of – Developing and supporting community institutions
programme
COMMUNITY TARGETING

needs significant staff time, skills and resources

+ Community members understand complex – Difficult to standardise or compare targeting
interacting causes of vulnerability and food between different communities, because decisions
insecurity better than outsiders. They can judge are based on subjective and relative judgement.
people‟s overall situation and needs without
– „Community‟ may not include the most vulnerable
complex algorithms to combine indicators.
groups, or in some situations (e.g. displacement,
conflict) community institutions may have broken
down.
– Costs to community decision-makers, in time and
trouble, can be high.
– Risk of bias, corruption, intimidation, theft or error –
requires monitoring / auditing (as with
administrative targeting)
 No corruption or bias in selection – beneficiaries – Needs good information and analysis to determine
themselves decide whether to participate which costs and benefits will select the intended
 No administrative costs of direct selection target group
– Programme must be able to accept everyone who
SELF-TARGETING

self-selects, otherwise competition for resources

may exclude the powerless and create
opportunities for bribery or pressure
– If resources are not enough to assist everyone who
self-selects, administrative or community targeting
methods will be needed in addition to self-targeting.
– Can result in high exclusion and inclusion errors
– Not very effective in relief and crisis situations
– Not effective with free distributions
+ „Impersonal‟ selection – no corruption or bias – Often benefits the better-off more than the poor
+ No administrative costs of direct selection (high inclusion errors)
MARKET-BASED

(information and management) – Can exclude the poor and vulnerable who lack
+ Can recoup part of programme costs exchange entitlements (high exclusion errors)
 Can avoid dependency and disruption to local – Needs good information and analysis of the market
economy (if well-managed) and the economic position of the target group
– Can displace private-sector traders and / or
discourage production (if poorly-managed)

29
TARGETING RELIEF FOOD ASSISTANCE PART THREE: TOOLBOX

13.5. Market-based targeting

In market-based targeting, there is no direct selection of beneficiaries. Instead, interventions are made in
the market to influence the price, supply or demand either of food or of commodities sold by the target
group in exchange for food. Examples are: releasing strategic grain stocks onto markets in targeted areas
(to increase the supply and therefore bring down or stabilize the price); or purchasing livestock during a
food crisis (to maintain demand and keep up the price, thus supporting the purchasing power of livestock
owners).
Market-based targeting should be based on a sound analysis of the economic position of the intended
target group. For example, general price subsidies may be applied selectively to staple foods eaten
mainly by the poor (in economic terms, ‘inferior commodities’ with ‘negative income elasticities of
demand’, meaning that their consumption falls as income rises). Livestock purchase schemes are based
on the analysis that pastoralists’ access to food is highly vulnerable to changes in the terms of trade
between animals and grain. Increasing the supply of grain to targeted markets assumes that the
vulnerable or food-insecure people served by those markets have the purchasing power to buy food, but
that rising prices threaten their access to an adequate diet.
Well-designed market interventions can be very effective in supporting livelihood systems, but their
household-level targeting impact can be quite indiscriminate. For example, food price subsidies often
benefit better-off people more than the poor, thus producing high inclusion errors, simply because richer
people buy more food. This effect can be counteracted by fine-tuning the commodities chosen, or by
selling through restricted outlets. Similarly, without additional targeting systems, “livestock price support
helps pastoralists proportionately to the size of their herds, with large owners gaining much more than
smaller ones” (Drèze and Sen 1989:109). Exclusion errors can also be high with market targeting,
because the target groups may lack the purchasing power or exchange entitlements needed to benefit
from the market intervention.

13.6. Combining methods

The great majority of targeted relief and social protection programmes do not use a single targeting
method, but combine elements of two or more methods to ensure successful targeting. In this way the
weaknesses or disadvantages of each method can be offset. For example, the PSNP uses a combination
of administrative, community, and self-targeting elements. Table 7 summarises some key advantages
and disadvantages of the main targeting methods which should be considered when choosing and
combining methods.

14. Targeting ‘errors’ and problems

14.1. Perfect (theoretical) targeting

This section uses a hypothetical, simplified situation to show how targeting would work in a perfect
world, and why targeting is more effective and efficient than an untargeted distribution. This is not to
suggest that perfect targeting is a realistic aim: it is rarely, if ever, achieved. The intention is to clarify the
logic of targeting, and to set out a framework for analysing the types of problem that arise in practice.
Figure 5 shows the situation at the time of an emergency food needs assessment. The hypothetical
community consists of ten households. For the sake of this illustration we assume that they are all the
same size (4 people) and therefore have the same monthly minimum food need (say 4 x 15 kg, = 60 kg,
shown as the survival threshold in the graph). 18 All have been affected by the same disaster. However,
their food access varies because of differences in their livelihoods, assets and coping capacity. Six of
them have a food gap, meaning that the food they can access through production, purchase and all other
means is less than their consumption needs, while the other four (households 7 to 10) are able to meet

18 The actual survival threshold values used in HEA analysis are more nuanced than this simplified example (see the
LIU Seasonal Assessment Manual).

30
TARGETING RELIEF FOOD ASSISTANCE PART THREE: TOOLBOX

their basic food needs without assistance. The total of the six food-insecure households’ gaps is 160kg,
so the needs assessment concludes that 160kg is the quantity of food assistance needed for the
community.

Figure 5: Hypothetical situation at needs assessment

100
Food Food gap
access 90
without 80
assistance 70
(a) 60 – (a) 60
hh 1 20 40 50
hh 2 25 35 40
hh 3 30 30 30
20
hh 4 35 25
10
hh 5 40 20 0
hh 6 50 10 hh 1 hh 2 hh 3 hh 4 hh 5 hh 6 hh 7 hh 8 hh 9 hh 10
hh 7 60 0
hh 8 70 0 Hh food access without assistance
hh 9 80 0 Survival threshold (minimum household food needs)
hh 10 90 0
Total 160 = aid allocated
Hh = household

Figure 6 illustrates the effect on each household’s food access if the 160kg is shared equally among the
whole community (that is, it is untargeted at the household level). Although the distribution is equal it is
not equitable (fair), because it does not take account of differences in need. In this example, spreading
the aid over all ten households produces a dilution effect. This means that the quantity received by the
target group is not enough to solve their acute food shortage: households 1 to 5 still cannot meet their
basic consumption needs, even though the community received the right amount of aid to help all those
with food gaps.

Figure 6: Effect of blanket (untargeted) distribution

Food Blanket Food access 120

access aid after blanket
100
without distributed distribution
assistance (a) + (b)
80
(a) (b)
hh 1 20 16 36 60
hh 2 25 16 41
40
hh 3 30 16 46
hh 4 35 16 51 20
hh 5 40 16 56
0
hh 6 50 16 66 hh 1 hh 2 hh 3 hh 4 hh 5 hh 6 hh 7 hh 8 hh 9 hh 10
hh 7 60 16 76
hh 8 70 16 86
Hh food access without assistance
hh 9 80 16 96
Blanket (equal) aid distribution
hh 10 90 16 106 Survival threshold (minimum household food needs)
Total 160

31
TARGETING RELIEF FOOD ASSISTANCE PART THREE: TOOLBOX

Even though there is no exclusion error in this situation (i.e. all the target group are beneficiaries), the
distribution has failed to solve their problem. On the other hand, households 7 to 10 have been included
in the distribution even though they did not need it. This is considered an inclusion error, and the
quantity of aid received by them is considered as leakage or wasted resources. These types of targeting
error are all discussed in more detail in Section 14 below.
By contrast, Figure 7 shows the effect on food access if the same quantity of aid is perfectly targeted,
meaning that the amount of aid received by each household exactly matches their food gap. All
households in the community now have enough food; therefore the targeted distribution is more
effective. None of the food assistance has been wasted or distributed to people who did not need it;
therefore it is more efficient (achieving a greater benefit or impact for the same cost). No-one has been
wrongly excluded, or wrongly included.

Figure 7: Effect of perfect targeting

Food Targeted Food access 100

access aid after 90
without distributed perfectly 80
assistance (= hh food targeted 70
gap) distribution 60
(a) (t) (a) + (t)
50
hh 1 20 40 60
40
hh 2 25 35 60
30
hh 3 30 30 60 20
hh 4 35 25 60 10
hh 5 40 20 60 0

hh 6 50 10 60 hh 1 hh 2 hh 3 hh 4 hh 5 hh 6 hh 7 hh 8 hh 9 hh 10

hh 7 60 0 60
Hh food access without assistance
hh 8 70 0 70
Perfectly targeted aid
hh 9 80 0 80
Survival threshold (minimum household food needs)
hh 10 90 0 90
Total 160

Of course, this illustration makes a number of assumptions that are unlikely to hold true in real life. For
example, it is extremely difficult to make such an accurate needs assessment; the total quantity of aid
distributed often does not match the total needed; and at household level, it would be extremely
complicated and resource-intensive to match aid entitlements exactly to food gaps. These are all
acknowledged challenges which are likely to affect targeting in practice. The task of everyone involved in
relief food assistance is to address these challenges in ways that make the targeting outcome (and
therefore the impact on people suffering from acute food insecurity) as close to the ideal as realistically
possible.
In assessing how good a real-world targeting system is, it is common to talk about ‘errors’ (particularly
inclusion and exclusion errors). In this context, ‘error’ is a technical term borrowed from statistics,
meaning roughly the same as ‘inaccuracy’. It does not necessarily mean that someone has made a
mistake: in fact, some degree of inaccuracy is inherent in most targeting systems, either because of the
way they are designed or because of constraints in implementation. Sometimes a certain level of
targeting error is accepted or even intended in order to achieve priority goals, or because the costs of
more accurate targeting are judged to outweigh its benefits.
There are two broad categories of inaccuracy or failure which can reduce the effectiveness of targeting:
problems with who receives the aid (‘benefit incidence’), and problems with how much aid is received by
each beneficiary (‘benefit levels’) (Devereux 2004). These are explained in the following sections.

32
TARGETING RELIEF FOOD ASSISTANCE PART THREE: TOOLBOX

14.2. ‘Benefit incidence’ errors: exclusion and inclusion

The accuracy of a targeting system is often measured in terms of how many people (or what percentage
of people) are correctly or incorrectly receiving assistance. Figure 8 represents what happens when the
target group and the beneficiaries (that is, the people who actually receive the benefits) 19 are not exactly
the same.

Figure 8: Exclusion and inclusion errors

Total population

a c
b Beneficiaries
Target
population (people receiving
assistance)

EXCLUSION ERROR CORRECT TARGETING INCLUSION ERROR

a / (a+b) Coverage = b / (a+b) c / (b+c)

Adapted from Taylor and Seaman 2004, Figure 1 (p.5)

 Exclusion error is the percentage of the eligible population or target group who do not receive
benefits. Exclusion can be caused by various factors such as under-resourcing; bias or discrimination
in the targeting process; lack of information or capacity; poor choice of criteria; or poor
communication.
 Inclusion error is the percentage of beneficiaries who are wrongly included (that is, they are not
eligible according to the targeting criteria, or they do not belong to the target group). This type of
error, too, can be caused by a number of factors including over-resourcing; bias or favouritism; social
pressures; poor choice of criteria; or disagreement with the principle of selective targeting.
 The green area (b), where the intended and actual beneficiaries overlap, represents correct targeting.
This can be measured either as a percentage of the target group, or as a percentage of the
beneficiaries. Usually the preferred indicator is coverage- the percentage of the target group who are
correctly identified and receive benefits.
 Coverage and exclusion error are complementary – in other words, if the coverage is 70% of the
target population, then the exclusion error must be 30% (100-70).

A perfect targeting system would identify all the target group (zero exclusion error and 100% coverage),
and only the target group (zero inclusion error). However, as mentioned above, some degree of
inaccuracy is inherent in most targeting systems. Strategic judgements must often be made about the
type(s) and degree of inaccuracy that are acceptable, and which type(s) of error it is most important to
minimise, given the situation and the objectives of the response. In emergency relief operations where

19 Confusion has sometimes arisen from the use of the term „beneficiaries‟ in the Humanitarian Requirements and
other documents to mean the intended beneficiary numbers (i.e. the target group or number of people in need). In
these guidelines, „beneficiary‟ means a person who has actually benefited (i.e. has received assistance).

33
TARGETING RELIEF FOOD ASSISTANCE PART THREE: TOOLBOX

lives and livelihoods may be at risk, priority is usually given to minimising exclusion errors (or, to put it
the other way round, maximising the coverage of the population in need).
At the design stage, there tends to be a trade-off between inclusion and exclusion errors. When strict
selection criteria and procedures are put in place to reduce inclusion errors, this usually increases
exclusion errors and vice versa. In implementation, the relationship is different: once the overall quantity
of aid has been allocated, high inclusion errors will either cause increased exclusion (if the household
ration is fixed), or dilution of benefits (if rations are reduced to cover a larger number of beneficiaries).
(Taylor and Seaman 2004)

14.3. ‘Benefit level’ errors: dilution and leakage

Dilution is a common targeting problem in which resources are shared among a larger number of people
than the programme was designed for, resulting in each beneficiary receiving an inadequate amount of
aid. The term ‘dilution’ is the opposite of ‘concentration’. It is a problem because a key purpose of
targeting is to concentrate resources where they will have the greatest impact on food insecurity. If
dilution is very high, the impact of the assistance programme can be significantly reduced.
Dilution can be caused by under-resourcing (when the total resources provided are not enough to deliver
the intended quantity of aid to all the eligible target group members). It also happens when communities
disagree with the principle of targeting, and consider it fairer for everyone to have a share. Well-
intentioned strategies to spread the available aid as widely as possible (such as rotation of beneficiary
lists or areas, capping the number of family members for household allocations, and excessive cutting of
the ration per person) can also cause dilution of the impact at beneficiary level.
 Dilution can be quantified by the average amount of aid received by each beneficiary, as a
percentage of the planned or standard ration.

Two other measures of targeting accuracy in terms of quantities of aid (as opposed to numbers of
people) are derived at a more aggregate level:
 Targeting effectiveness is the percentage of the total quantity of aid provided which is actually
received by the target group.20 Thus it is a summary measure of how successful the targeting system
has been in ensuring the available assistance reaches the right people. This can be assessed either
for a defined area or community, or for the whole of a relief operation (depending on the purpose of
the analysis and the data available).
 Leakage is the percentage of the aid provided which is not received by the target group. Leaked
resources may have gone to ‘inclusion error’ beneficiaries (i.e. people who did not need the
assistance), or they may have been diverted out of the relief distribution system.
 Effectiveness and leakage are complementary – that is, if the effectiveness is 80%, then the leakage
must be 20% (100-80).
To illustrate this from the hypothetical example above (Figure 6), the total aid received by target-group
members (households 1 to 6) was 96kg, and the total aid delivered to the community was 160kg.
Therefore, the targeting effectiveness in this case was 60% (96/160). Conversely, the leakage was 40%
(100-60, or 64/160).

14.4. Timing errors

Targeting errors can also arise when people receive food assistance at the wrong time (nearly always
later than needed), or for a longer or shorter period than necessary to meet their relief and recovery
needs. When aid distributions are late, they may fail in the fundamental goal of saving lives and
livelihoods. If delayed until the next harvest season, they can also cause income losses and disincentive

20 This useful definition of „targeting effectiveness‟ is from the World Bank, but note that the term is sometimes used
with different meanings by other agencies.

34
TARGETING RELIEF FOOD ASSISTANCE PART THREE: TOOLBOX

effects for local producers by depressing market demand and prices. Emergency relief assistance (in
contrast to safety-net and developmental transfers) should always be planned as a temporary, short-
term operation with a clearly defined start and end date objectively based on an analysis of the disaster
impact and the seasonality of local food and livelihood systems.

14.5. Causes of targeting error

There are three broad sources of inaccuracy in targeting: technical or design problems, implementation
problems, and resource problems. It is important to understand why a targeting error is arising, in order
to take the right action to improve it.
Technical or design problems may be due to:
 Inaccurate needs or vulnerability assessment;
 Gaps in data or information;
 Inappropriate or impractical criteria;
 Poor choice of targeting method (e.g. using self-targeting in a situation where screening criteria are
needed); or
 Poor analysis of power structures and interest groups, leaving the targeting open to ‘elite capture’.
Implementation problems can include:
 Weak administrative capacity;
 Poor accountability (lack or weakness of audit and appeal procedures);
 Poor communication or consultation (potential beneficiaries are not informed about the targeting or
the relief operation in general);
 Clash of values (e.g. the community’s concepts of vulnerability and fairness do not match the target
group definitions);
 Malpractice: diversion or misuse of resources; or
 Competing objectives: relief resources may be used to support other local agendas or programmes.
Resource problems are basically of two kinds, which arise when the assistance provided is either less or
more than the actual need.
 Under-resourcing or late delivery of assistance is likely to cause exclusion errors and/or dilution. Local
decision-makers often have to make very difficult choices about the best way to distribute smaller
quantities of aid than they had planned for, especially if community targeting lists have already been
compiled.
 Over-resourcing, on the other hand, is likely to cause inclusion errors and may foster a sense of
entitlement or dependency among habitual beneficiaries.

15. Tools for monitoring and evaluation

15.1. Types of monitoring
Targeting can be monitored in a variety of ways, from a statistically representative questionnaire survey
or systematic stakeholder consultations to an informal conversation with beneficiaries met by chance at a
distribution site. A dedicated mission or survey may be launched to investigate targeting issues, or more
often questions on targeting may be added to routine monitoring checklists. Since targeting is a theme
which runs through every stage of the disaster response, issues related to targeting can also be
monitored at each stage. Table 8 (from the Emergency Nutrition Network’s special supplement on relief
targeting) summarises the key questions and different instruments that can generate information and
feedback on targeting during the needs assessment, beneficiary selection, and distribution phases

35
TARGETING RELIEF FOOD ASSISTANCE PART THREE: TOOLBOX

Table 8: Key questions and methods for monitoring relief targeting

a) Monitoring targeting issues at different stages of the relief operation

b) Characteristics of the monitoring methods listed in Table a.

Source: Taylor and Seaman 2004:22-3, Tables 2 and 3

Whichever approaches are used, the overall monitoring system should track both the targeting process
(how well is the targeting system working?), and its outcomes (who received what, and did it meet their
needs?). In terms of the purpose and use of information, monitoring can be corrective (i.e. ‘action-
based’, for an immediate response to any problems detected) or more analytical, for a more
comprehensive and representative understanding of how the whole system has performed. Analytical
monitoring should also feed back into improvements, but in the longer term. An effective monitoring
system should use a combination of all these elements, in order to verify that adequate assistance is
reaching the right people and that the process is fair and transparent.

36
TARGETING RELIEF FOOD ASSISTANCE PART THREE: TOOLBOX

SPHERE Key Indicator 4:

“Distribution systems are monitored to ensure that targeting criteria are respected and that
timely corrective action is taken when necessary.”
(see Box 3 in Part Two)

15.2. Monitoring the process (how well is the targeting system working?)
Monitoring how a targeting system is functioning involves checking how decisions are made and who by,
whether the decisions are timely and fair, and whether the process is transparent and accountable.
Problems should be detected and corrected as soon as possible. It is equally important to investigate
why problems are happening (see Section 14.5 above for a checklist of possible causes of targeting error),
in order to recommend what should be done to correct them. Table 9 gives a provisional list of questions
for monitoring the process, especially at community level.

DRM Principle 13: Avoid malpractice

(see Table 3 in Part Two)

15.3. Monitoring the outcome (who received what?)

Assessing the outcome of targeting – that is, verifying whether the assistance reached the right people, in
the right quantities, to meet their emergency food needs – is, of course, crucial. Table 10 suggests some
quantitative and qualitative indicators under four key questions:
 How many people were correctly or incorrectly targeted?
 Who is included and excluded?
 How much assistance was received by the target group? and
 How well did the assistance received meet their needs?

37
TARGETING RELIEF FOOD ASSISTANCE PART THREE: TOOLBOX

Table 9: A basic checklist for monitoring the targeting process

Transparency and information flows:

 Do members of the targeted community (beneficiaries and non-beneficiaries) know:

 the criteria used to select relief beneficiaries in this community?
 the quantity of assistance allocated per person, and the timing of distributions?
 who to complain to if there is any problem with the targeting?
 where to see the registration list? (Was it posted in a public place, and/or read out in a full community
meeting?)
 Do community representatives (relief committee or others managing the targeting) know:
 the general criteria and principles for relief targeting?
 the distribution plan for their community?
 the main findings of the needs assessment for their community?

Participation:
 Who decided the registration list? (e.g. full community meeting; community committee or representatives;
kebele council)
 What was the gender composition of the people who decided? (all or mostly men? Equal men and
women? All or mostly women?)
 Who decided the criteria for relief eligibility? (e.g. full community meeting; community committee or
representatives; kebele council)
 Do community members think the criteria are right? (i.e. do the criteria match the most vulnerable / needy
people in this community?)
 Do community members think the targeting process was fair? (If not, why not?)

Accountability and safeguards:

Information sources:
 Discuss with the appeals body (or whoever is responsible for hearing appeals if there is no formal body);
 Check / audit records of appeals and action taken (if there are no records, why not – is support or capacity
development needed?).
 Talk to complainants separately; ask their individual stories and views.
 Discuss with focus groups if possible (e.g. groups of beneficiaries; groups of people who appealed).
 OR include these issues in a household questionnaire.
Try to find out:
 Have there been any appeals or complaints about the targeting process?
 What kinds of appeal or complaint? (e.g. exclusion; under-registration of household members; wrong
inclusion of others; corruption?) If possible, estimate number of appeals of each type.
 What process did the appellants / complainants follow (who did they go to, what happened etc.)?
 What action or decision was taken in each case?

38
TARGETING RELIEF FOOD ASSISTANCE PART THREE: TOOLBOX

Table 10: Questions and indicators for monitoring targeting outcomes

Quantitative indicators Qualitative indicators / sources

1. How many people were correctly or incorrectly identified as beneficiaries?

 Coverage = % of target group receiving benefits  Community estimates (consult targeting

 Exclusion = % of target group not receiving committee, focus groups, &/or key informants).
benefits Participatory methods such as proportional piling
can give an estimate of numbers.
 Inclusion = % of beneficiaries who do not meet the  Monitor‟s observations
eligibility criteria
 Audit of appeal records

2. Who is included, who is excluded?

 Statistical comparison of beneficiary & non-  Compare case-studies or separate focus groups
beneficiary characteristics21, e.g. of beneficiaries and non-beneficiaries.
 Coping strategies  Profile or interview people identified as wrongly
 Asset levels included or excluded – are they:
 Dependency ratios  Gross errors (i.e. neediest / most vulnerable
 Diet / food access people excluded; well-off or powerful included),
 Other characteristics relevant to local targeting or
criteria  Marginal errors (i.e. people just inside or
outside eligibility threshold, where targeting
decision could be difficult or disputed)?
 Do wrongly excluded or included groups have
any factor in common?
 Investigate why inclusion / exclusion errors are
happening (e.g. resource shortage? corruption?
poor understanding of guidelines?)

3. How much of the resources for distribution reached the right people?

 Effectiveness = % of delivered food assistance  Audit of registration lists & ration cards.
received by the target group  Interview beneficiaries – check quantities received,
 Leakage = % of delivered food assistance not number of household members registered.
received by the target group
 Dilution=average quantity of aid received per
person, compared to planned ration

4. What was the impact (did the assistance meet people’s needs)?

 Assistance received as % of household food gap  Interview beneficiaries – did the type and quantity
of assistance meet their needs?

21 When comparing beneficiaries and non-beneficiaries after distributions, it must be remembered that the
beneficiaries‟ situation has (we hope) been improved by the assistance they received. This should be kept in mind
when interpreting the comparison.

39
TARGETING RELIEF FOOD ASSISTANCE PART FOUR: NATIONAL TARGETING SYSTEM

PART FOUR: ETHIOPIA’S NATIONAL TARGETING SYSTEM

16. Targeting throughout the disaster-response cycle

As we have seen in Part 3, for an emergency intervention to reach the right people, with the right kind
and quantity of aid, in the right place and at the right time, it is not enough to have a method of
registering eligible households or individuals. Early warning, needs assessment, planning and logistics
must also be effective and co-ordinated with each other. The system for identifying and registering
beneficiaries must match the general definition of the target groups generated by the early warning and
assessment process (that is, the logic of who has been affected, in what ways, by the specified disaster
event). Once beneficiaries have been registered, the distribution process must also ensure that those are
the people who actually receive the assistance – that is, that the relief reaches the target groups. Finally,
there must be a system to verify or check the targeting outcomes, and to use that information to make
any improvements needed. In operational terms, therefore, successful targeting depends on
interconnected actions at each stage of the disaster-response cycle (illustrated in Figure 9 below).

Figure 9: The chain of actions needed to ensure good targeting

EARLY WARNING:
identify emergency,
analyse risks and
MONITORING: vulnerabilities ASSESSMENT:
verify targeting, define target group,
correct problems estimate numbers
and locations

DISTRIBUTION: PLANNING:
ensure aid reaches choose modalities,
the target group rations, distribution
members SELECTION schedule
of target group
members: identify
beneficiaries

In view of this need for connectedness, the following sections outline how the national system for
targeting relief food assistance currently works, in order to contextualise the more detailed guidance on
beneficiary-level targeting which is given in Part 5. Various elements of this system are expected to
change or evolve over the coming few years, as early warning and relief management are increasingly
decentralised. The core principles and the logic of ‘joined-up targeting’ set out in these guidelines will
remain applicable, and should be considered whenever institutional or procedural changes are made
which may affect the targeting of relief assistance.

40
 TARGETING RELIEF FOOD ASSISTANCE PART FOUR: NATIONAL TARGETING SYSTEM

Figure 10: Overview of the relief targeting system, from federal to beneficiary level

INSTITUTIONAL CHAIN: INFORMATION

FUNCTIONS
DECISION-MAKING & ACCOUNTABILITY SOURCES
Identify areas at risk
Early Warning Predict disasters
Federal EWRD LIU System Provide info for needs &
GEOGRAPHICAL TARGETING ================================================

DRMFSS Continuous (monthly) verification assessments

Humanitarian Generate rolling annual

Seasonal Needs planning figures
partners Assessments (locations and numbers
Meher; belg, etc.
in need)
------ LEVEL 1 ------

DRM Technical
Working Group Prioritise areas of need
“Hot-spot” when resources are less
Analysis than planned; identify
as needed
Prioritisation nutritional crises
Committee
Confirm / adjust actual
Regional Verification needs per wereda,
EWR / FS body* Assessments according to changing
as needed
conditions

Assess sudden or
Disaster area unexpected relief needs
Assessments (outside planning figures)
as needed

Wereda roles:
Wereda experts Target kebeles /
and monitoring

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ BENEFICIARY TARGETING

allocate
information resources within
------ LEVEL 2 ------

wereda
Wereda
WFP Sub-offices Provide data for
and Field EWS and
Monitors assessments

NGO monitoring Supervise

and field staff beneficiary
targeting

HEA baselines
Kebele
& impact / needs
assessment
------ LEVEL 3 ------

Community Nutritional Identify

(village / clan) Potential channels for beneficiaries
Committee or leaders appeals or complaints screening
(households /
individuals)

Community Community
members knowledge
& local EW

* Regional structures and institutional names vary.

41
TARGETING RELIEF FOOD ASSISTANCE PART FOUR: NATIONAL TARGETING SYSTEM

17. Targeting levels, institutions and roles

Figure 10 gives a simplified overview of the institutions, information sources, and levels of targeting
involved in the existing national system for targeting relief food assistance. A quick glance conveys that
there are many organisations and levels of government involved in generating information and making
decisions on targeting. However, not all of these are necessarily involved in a given relief operation: this
will depend on the type of emergency and the most efficient way to assess and respond to it.
The system can be divided into three levels, labelled on the left of the diagram: Level 1 represents
geographical targeting from federal to wereda level, Level 2 is geographical targeting within the wereda
(to kebeles and communities)22, and Level 3 is the targeting of beneficiary households, groups or
individuals.

17.1. Level 1: Geographical targeting from federal to wereda level

Level 1 decisions on geographical targeting - that is, the identification of areas in need of relief food
assistance, the estimation of numbers in need and quantities of food required, and the allocation of
resources to regions and weredas - are made at the national and regional levels. Information for these
decisions comes from various sources, including regular early warning reports from the weredas,
seasonal needs assessments or disaster area assessments (for fast-onset disasters). DRMFSS plans to
strengthen the regular early warning system and phase out or reduce the importance of the seasonal
assessments, over the next few years.

17.1.a. Needs assessments

The starting point of the emergency response process is the systematic assessment of disaster impacts
and needs. In Ethiopia, a major multi-agency needs assessment is conducted before the main (meher)
national harvest, in November / December. This assessment covers both crop-farming and pastoralist
parts of the country, and generates planning figures for expected relief needs in the coming year (from
January to December). The figures are updated later in the year, by a second assessment in May/ June
timed to take account of the secondary (belg) harvest and the gu rains in the southern pastoralist areas.
Mid-season assessments are conducted to give early predictions of food security conditions to assist with
forward planning and pre-positioning of resources. Further assessments may be added at different times
in the year as needed.
The methodology for needs assessment has been significantly changed since the previous targeting
guidelines were issued. The national system now uses the multi-hazard, livelihoods-based Household
Economy Approach (HEA) to make a holistic assessment of the impact of disaster events on household
food access. Technical details about the application of this approach to needs assessment can be found
in the LIU’s Seasonal Assessment Manual, while Section 18 below outlines how HEA methods and data
can be used for continuing improvements in targeting at different levels.

 See LIU 2009, Seasonal Assessment Manual

The institutionalisation of HEA in the nationwide system of needs assessment is an on-going process.
Although great progress has been made, ‘traditional methods’ of needs assessment (relying more heavily
on the food production and food-balance sheet approach, combined with local expertise and informed
judgement) are still used in some cases, either instead of HEA or combined with it. In the long term, it is
expected that the analytical framework and data-processing methods of HEA will be used throughout the

22 The kebele or PA (sub-district) is the lowest level of professional government administration, and the geographical
unit by which the boundaries of livelihood zones are defined. In general, kebeles are too big in population terms to
function as coherent communities where everyone knows each other. Therefore, for community-based targeting,
a further level of disaggregation is needed to social units such as the village or clan (see Section 19).

42
TARGETING RELIEF FOOD ASSISTANCE PART FOUR: NATIONAL TARGETING SYSTEM

system. Whichever methods are used, seasonal and ad-hoc needs assessments should broadly define the
following parameters for targeting:
 The locations, numbers and general characteristics (e.g. livelihood profile, vulnerability or poverty
ranking, demographic profiles) of people in need of relief;
 The type and quantity of assistance needed (food and non-food; general rations and/or
supplementary food; ration rate; or at Livelihood Zone level, the percentage of food needs unmet
and the size of the expected food gap); and
 The duration and timing of need (how many months of assistance are needed, and when in relation
to the seasonal calendar).

17.1.b. PSNP and relief needs

The needs assessment process generates estimates of the total expected shortfall in food access, and the
number of people expected to need assistance, for each wereda. In chronically food insecure weredas
where the PSNP is operational, part of this total shortfall will be met by resources already planned for
registered safety-net beneficiaries (predictable resource transfers for predictable levels of food gap).
Therefore, the number of people who will be supported by the PSNP is deducted from the total needs
estimate to arrive at the residual number who will need relief assistance for the temporary emergency
period.
The PSNP is also designed to scale up quickly and temporarily in response to transitory food insecurity in
its operational weredas, through its Contingency and Risk Financing (RF) budgets. Relief food assistance
should only be mobilised for PSNP weredas when the emergency needs exceed the quantity or type of
assistance the safety net can provide.23 This may arise, for example, if the scale or duration of needs is
greater than the resources that can be mobilised from the PSNP budgets; or if food commodities are
urgently needed while the PSNP can provide only cash; or if for whatever reason the Contingency and
Risk Financing mechanisms fail to work as planned. Guidelines for the coordination of relief and PSNP
targeting at household level are given in Part 5. Procedures for harmonising the emergency relief and
PSNP targeting systems, particularly with RF which was introduced in 2010, are still in development. The
implementation of these guidelines should therefore be managed so as to provide systematic feedback
and suggested improvements to these procedures, especially at the wereda level.

 See MoARD 2009, Final Guidelines for the PSNP Risk Financing Mechanism in Ethiopia

17.1.c. Information flows

The outcome of the needs assessment process is published in summary form in Humanitarian
Requirements Documents, issued jointly by the Ethiopian government and its humanitarian partners.
The main document, summarising food needs for twelve months and non-food needs for six months, is
issued at the beginning of the European calendar year, with periodic updates later in the year as needed.
These documents form a vital basis for the planning and resourcing of both food and non-food relief
assistance. In the past they have presented only the aggregate number of targeted beneficiaries per
region. From April 2011, a more detailed breakdown is given by wereda, showing numbers of targeted
beneficiaries, duration of assistance needed (number of months), and timing (the month when
distributions should start) (DRMFSS 2011). This additional detail is highly useful for the targeting process.

23 This applies specifically to food assistance and not to non-food emergency needs, which are not addressed by the
PSNP.

43
TARGETING RELIEF FOOD ASSISTANCE PART FOUR: NATIONAL TARGETING SYSTEM

Ideally, the HRD and a Risk Financing Requirements Document (RFRD) should be released at the same
time. The total number of people in need of transitory assistance would then be identified during the
needs assessment, and the RFRD would identify how many people in PSNP weredas are receiving support
from the Contingency Budget and the RFM. However, the procedures for implementing the RFM, and for
co-ordinating it with humanitarian relief, are still in development.

In order to further facilitate ‘joined-up targeting’, to improve the analytical linkages and feedback
between geographical and beneficiary-level targeting, to fine-tune the temporal targeting (timing and
duration of relief operations) in different parts of the country, and to enhance the accountability and
transparency of the whole targeting system, the analytical reports and data which inform the HRD (and
proposed RFRD) summaries should in future be shared with the relevant government departments and
humanitarian partners at federal and regional levels.

17.1.d. Regional verification

The planning figures generated by the seasonal needs assessments do not constitute a fixed distribution
plan for relief (unlike the PSNP, which is planned on a multi-year predictable resource basis). After the
first round of distribution, there is an on-going process of checking and adjusting the targeting plan
according to changing food security conditions (whether improving or deteriorating). The information on
which such adjustments are based comes primarily from the Regions, drawing on regular early warning
information from the weredas, supplemented by verification assessments or disaster area assessments as
needed. There are also less formal ‘feedback loops’ by which information on changing or unexpected
relief needs can be relayed to the regional or federal authorities so that they can respond accordingly.
These channels include verbal or written reporting from wereda authorities, and from humanitarian
partners in areas where they have a grassroots presence or are involved in relief distributions. These
sources may also report problems with targeting or distribution which require action from higher
authorities.
Technical guidelines for verification assessments were produced by the then DPPA in 2007, and were
updated following the devolution of responsibility for verification to the Regions.

 See DPPA 2007, Implementing Strategy for Verification Assessment

During the verification process the Regions are required to prioritise the areas in most urgent need, and
are authorised to re-allocate resources among weredas if necessary, within the overall total resources
agreed for their Region. For pastoralist zones, Regions may adapt this re-allocation facility to ensure that
relief is correctly targeted to mobile populations who do not stay in one wereda.
The Federal DRMFSS requires Regional verification of beneficiary numbers and needs in order to release
resources for distribution rounds. This process is intended to ensure proper targeting of assistance in
response to changing needs. It should not cause delays in the distribution process for people in need of
food assistance. All Regions should review and if necessary streamline their verification process, in
consultation with the Federal DRMFSS and relevant humanitarian partners, in order to ensure that the
verification requirement does not impede the timeliness of distributions and targeting.
Regions should also ensure that any changes in relief allocation among weredas are consistent with the
fundamental principles of targeting, that is they should be based on an objective assessment of needs
and prioritisation of the neediest. Weredas should not be rotated in order to spread resources: this is
likely to result in significant dilution of benefits for people in need of assistance. Information about any
changes in planned allocations to weredas, including the reasons for the changes, should be clearly

44
TARGETING RELIEF FOOD ASSISTANCE PART FOUR: NATIONAL TARGETING SYSTEM

communicated to local government and partners in advance. Unexpected cuts or delays in the relief
assistance received by weredas can also cause serious problems with beneficiary-level targeting (see Part
5). Relief distributions within a wereda or community should not be stopped without a phase-out plan,
including reasonable notice to beneficiaries and implementing staff. If the Region considers that the total
quantities of relief assistance allocated to it are insufficient they should appeal to the federal DRMFSS,
presenting evidence of the additional need or requesting a disaster area assessment if needed.

17.1.e. ‘Hot-spot’ Prioritisation

Adjustments to the distribution and targeting plan are also necessary when the resources available are
less than planned. This is a very common situation in relief operations, and it is often the major reason
for re-targeting. At wereda and community level (Levels 2 and 3 in the diagram), decisions are also
frequently needed about how to target the quantities of relief actually received when these are less than
the assessed needs or the planned distribution. Thus, an important but difficult feature of any relief
targeting system is the balancing of needs-based planning with resource-based prioritisation. The
targeting system should always aim to meet all the assessed relief needs: however, when this is not
possible priority must be given to the areas and people in greatest and most urgent need.

NEEDS-BASED RESOURCE-BASED
PLANNING PRIORITISATION

At national level, a Prioritisation Committee meets monthly to review the national relief food pipeline
situation in relation to the planned distributions. The distribution plan is compared with the available
stocks in country at the likely time of despatch, and if there is a shortfall the committee recommends to
DRMFSS which if any geographical areas should be prioritised, according to the latest malnutrition and
food security indicators. ‘Hot-spots’ are ranked by the Emergency Nutrition Coordination Unit (ENCU) of
DRMFSS on the basis of nutritional assessments, and this classification is supplemented by a range of
food security indicators and data analysed jointly by DRMFSS, WFP and FEWSNET. If necessary the
committee recommends a reduction in ration rates, either across the board (to spread the resource
shortfall evenly) or in selected areas (to maintain full rations to areas ranked 1 on the ‘hot-spot’
classification).

 See DRMFSS 2010a, Guidelines for Identification and Analysis of Areas Affected by Hazard
(Draft, to be finalised)

17.2. Level 2: Geographical targeting from wereda to community

Once the agreed number of beneficiaries and the quantity of aid have been allocated to the wereda level,
further targeting decisions must be made about dividing resources among the kebeles (PAs). In the past
these decisions were left largely to the wereda authorities with little or no guidance on how to proceed,
with the result that practice varied widely. This section aims to clarify wereda responsibilities on this
subject.
The allocation of beneficiary numbers and relief quantities to kebeles and communities within the
wereda must be based on the core principles which run through the whole targeting system from the
federal to the household level: firstly, that targeting should be based on an impartial assessment of the
needs of all groups, and secondly, that priority should be given to the most vulnerable and the most

45
TARGETING RELIEF FOOD ASSISTANCE PART FOUR: NATIONAL TARGETING SYSTEM

needy. Therefore it is not acceptable to share the wereda’s allocation of relief food assistance equally
among all kebeles (proportional to population or not), unless there is genuine and verifiable evidence
that the whole wereda is equally in need. Kebeles should be included or excluded according to the local
disaster impacts, and the quantity of assistance allocated to each included kebele should depend on the
numbers in need and the severity of disaster impacts.
To facilitate these decisions, each wereda should be provided with a copy of the detailed needs
assessment report for their area as soon as it has been approved by the appropriate Regional and Federal
authorities. The HEA-based impact assessment for the wereda should normally be the basis for targeting
kebeles. Livelihood zones are delineated by kebele boundaries, and lists of kebeles by wereda and
livelihood zone are available to assist with geographical targeting to below wereda level. The results of
the HEA Livelihoods Impact Analysis Sheet (LIAS) for the wereda, showing the estimated population in
need by livelihood zone and the rationale for the findings, should be made available for this purpose.
If there is disagreement with the needs analysis, or if the situation has changed since the assessment, or
if the needs assessment findings are not available to the wereda, then evidence for a needs-based
allocation of resources among the kebeles may be produced by documented wereda assessment
missions, reported to the Regions; or by analysis of existing early warning data; or by joint assessment
with Regional government or humanitarian partner staff when available. Each Region should decide how
to regulate and monitor local geographical targeting by the weredas.
For pastoralist zones, geographical targeting at wereda or kebele level may be inappropriate because of
population mobility. In these cases, the Regions may develop alternative means to identify and prioritise
the population groups who are most in need of assistance (e.g. by clan structures or catchment areas for
food distribution points).

17.3. Level 3: Beneficiary targeting

The last and most detailed level is beneficiary targeting – that is, the selection and registration of eligible
households and individuals. This is carried out at community level, overseen and supported by local
government and humanitarian partner organisations.
The wereda is a pivotal point of the whole system, as this is where geographical and beneficiary targeting
inter-connect. The wereda administration is in most cases responsible for generating regular early
warning reports, providing information for needs assessment missions, managing local geographical
targeting (that is, allocating resources among the kebeles), and supervising the process of beneficiary
targeting within kebeles and communities. Appeals and complaints about the beneficiary targeting are
often brought to the wereda if they cannot be resolved at community or kebele level. Wereda capacity,
in terms of material resources, staff numbers and expertise, is therefore a critical factor in effective relief
targeting. Humanitarian partners can make a major contribution at this level (see Part Five, which
focuses on this key level of targeting).

18. The Household Economy Approach applied to targeting 24

18.1. Assessing multi-hazard impacts and coping capacity

The Household Economy Approach (HEA) provides a logical framework for identifying areas, numbers and
types of people in need of assistance because of disaster impacts on their livelihoods from any type of
hazard (or hazards - either singly, or in combination). It has been applied to needs assessment and
geographical targeting in Ethiopia since 2006, when the LIU (Livelihoods Integration Unit) was established
in the DRMFSS.
The LIU uses the HEA to assess the emergency needs resulting from hazards such as drought (loss of crop
and livestock production) and market price change. The analysis, summarised in Figure 11, involves

24 This section incorporates material provided by Mark Lawrence (FEG) and Zerihun Mekuria (LIU).

46
TARGETING RELIEF FOOD ASSISTANCE PART FOUR: NATIONAL TARGETING SYSTEM

combining three types of types of information: baseline information on livelihoods (i.e. how people in a
given area normally access food and income), information on hazard (i.e. factors affecting access to food
and income, such as crop production or market prices) and information on coping strategies (i.e. the
sources of food and income that people turn to when exposed to a hazard).

Figure 11: Methodological Framework: the Household Economy Approach (HEA)

The types of coping strategy included in the analysis are important: some examples are given in Table 11,
although the range of coping strategies actually available to people, and the strategies that are
considered high cost or unacceptable, vary by livelihood zone and wealth group. Livelihood zones (LZs)
are “geographical areas within which people share broadly the same patterns of access to food and
income and the same access to markets, thus making them vulnerable to the same hazards” (Boudreau
2009: v).
‘Low cost’ and ‘medium cost’ coping strategies are acceptable and sustainable ways for households to
reduce their food access deficit in a period of acute food insecurity, without compromising their recovery
prospects and their future livelihoods. They are therefore included in the outcome analysis: if
households can meet their needs using these strategies, they are not considered in need of relief
assistance.
‘High cost’ strategies, also sometimes called survival or distress strategies, are those which are damaging
to people’s health or safety; those which risk leaving the household destitute or chronically food insecure
and unable to recover their livelihoods; those which are damaging to the environment or natural
resources; and those considered socially unacceptable or shameful. These strategies are not included in
the outcome analysis. Households should not be forced to employ these strategies to meet their food
needs, and should be provided with relief assistance to protect them from doing so.
Note that Table 11 shows generalised examples of coping strategies which might fall into the different
categories. Local conditions and social attitudes will determine exactly which strategies are acceptable or
damaging in a given situation. For targeting, the classification of coping strategies should therefore be
locally determined and should form part of the consultation with community representatives (see Section
21 below). Damaging strategies are defined in general as “strategies that undermine future means of
livelihood, dignity or nutritional health, increase long-term vulnerability, or are illegal or not socially
acceptable.” (WFP EFSA)

47
TARGETING RELIEF FOOD ASSISTANCE PART FOUR: NATIONAL TARGETING SYSTEM

Table 11: Types of Coping Strategy in HEA Analysis

Low-Cost (included in outcome analysis)

 Reduced expenditure on non-essential items (beer, cigarettes, ceremonies, festivals, expensive clothing, meat, sugar,
more expensive staples etc).
 Harvesting of reserve crops (e.g. enset, cassava)
 Consumption rather than sale of any crop surplus
Medium Cost (included in outcome analysis)
 Increased sale / slaughter of livestock (sustainable)
 Intensification of local labour activities
 Short-term / seasonal labour migration
 Intensification of self-employment activities
 Increased remittance income
 Increased social support / gifts
 Borrowing of food / cash
 Sale of non-productive assets (jewellery, clothing etc.)
 Collection of wild foods
High-Cost (excluded from outcome analysis)
 Unsustainable sale / slaughter of livestock
 Long-term / permanent migration (including distress migration of whole households)
 Excessive (environmentally damaging) sale of firewood or charcoal
 Sale or mortgaging of productive assets (land, tools, seeds, etc.)
 Prostitution
 Reduced expenditure on productive inputs (fertilizer, livestock drugs etc.)
 Reduced expenditure on health and education
 Reduced expenditure on water
 Decreased food intake

18.2. Geographical targeting by livelihood zone

The HEA analysis estimates the number of people facing a deficit, and the duration of assistance required,
by wereda and by livelihood zone. This information can be used to target assistance down to the level of
the livelihood zone within each wereda. For example, Figure 12 shows the food security situation of
people in different parts of one wereda (Angacha) in 2006.

Although the total grain production for the wereda was up that year, the increase was due to good
harvests of wheat and barley in the highlands and masked a severe drop in the production of maize, a
major crop in the lowland part of the wereda. Consequently, poor households in kebeles belonging to
the Badawacho-Alaba Maize LZ were unable to meet their minimum food needs (survival deficit shown in
red on the right-hand graph), while poor households in the nearby Hadiya-Kembata Cereal and Enset LZ
(shown on the left) experienced no problem.

18.3. Seasonality: timing and duration of relief needs

The LIU / HEA baselines contain a wealth of information about the peak months of annual food shortage
(‘hungry seasons’) in different livelihood zones, and the impact analysis can predict the level of deficit in
each LZ by month. More use should be made of this information to plan the best timing of seasonal
assessments, and the timing and duration of relief distributions, in each area. This information could also
help in prioritising areas for different distribution rounds during the year.
The planned decentralisation of relief management would facilitate better seasonal and temporal
targeting within different Regions in the future. In the meantime, an analytical annex to the national

48
TARGETING RELIEF FOOD ASSISTANCE PART FOUR: NATIONAL TARGETING SYSTEM

Humanitarian Requirements Document setting out the expected needs by location and month would
greatly improve this aspect of targeting.

Figure 12: Targeting by Livelihood Zone - different outcomes between LZs within one wereda

18.4. Household targeting: setting provisional criteria

Figure 13 shows how different wealth groups in the same livelihood zone can be affected in very different
ways by the same set of hazards, not only because of their asset levels but because their livelihood
profiles are different: they rely on different sources of food and cash income, and are therefore affected
differently when a hazard affects those sources.
The HEA analysis identifies which wealth groups in each LZ are facing a deficit (frequently just the ‘very
poor’, but often the ‘very poor’ and the ‘poor’; and occasionally the ‘middle’ and ‘better-off’). Wealth
breakdown tables for each livelihood zone give the main characteristics of each wealth group (in terms of
landholding, livestock and other assets). The baselines also describe the livelihood profiles of groups
expected to be worst affected by the particular hazards currently affecting the wereda. These
characteristics can provide guidance to wereda officials in setting provisional household targeting criteria
for discussion with communities (see Part Five).
Separate analyses are run for each wealth group in each livelihood zone within each wereda. Since there
is an average of two livelihood zones per wereda and four wealth groups per livelihood zone, this means
that the analysis is able to discriminate between the needs of – on average – eight different types of
household within each wereda.

49
TARGETING RELIEF FOOD ASSISTANCE PART FOUR: NATIONAL TARGETING SYSTEM

Figure 13: Targeting by Wealth Group - different outcomes within one livelihood zone

A full set of wereda posters, tailored to the specific characteristics of each wereda, has been prepared
and distributed to all weredas in Ethiopia as part of the LIU project. Each poster includes a map of the
wereda showing which kebeles are in which livelihood zone, plus wealth-group tables, seasonal calendars
and seasonal consumption patterns for each livelihood zone in the wereda.

50
TARGETING RELIEF FOOD ASSISTANCE PART FIVE: BENEFICIARY TARGETING

PART FIVE: BENEFICIARY TARGETING

19. Combined community and administrative targeting

At the level of distribution to households and individuals, responsibility for identifying the right
beneficiaries of relief food assistance rests with the community, supported and guided by local
government (wereda and kebele), in co-operation with humanitarian partners wherever such partners
are involved in relief operations and have staff available in the field. This approach of community
participation in targeting combined with administrative guidance and oversight has been widely used in
international relief operations since the 1990s. It has been in use in Ethiopia for several decades, and
was formalised by the previous Food Aid Targeting Guidelines in 2000. It has also been adopted and
developed by the PSNP.
As outlined in Part Three (Toolbox), the participation of beneficiary communities in targeting has a
number of advantages. Firstly, community members have much better ‘inside information’ about the
complex interaction of vulnerability factors within their own livelihood and social systems: therefore they
can make judgements about the overall food security situation of their neighbours, without employing
the costly, time-consuming and potentially inaccurate process of data collection and analysis that would
be needed for ‘outsiders’ (administrators or agency staff) to identify the neediest households without
community help. Secondly, targeting criteria are more likely to be adhered to in practice if the
community has been consulted and agrees with them, and has had the opportunity to adapt them to
local realities. Thirdly, effective participation in the targeting and distribution process should also
enhance accountability, by providing community members with information and understanding about
the purpose and rules of the relief operation, and about the distribution plans: this information makes it
possible for them to challenge poor decisions and hold decision-makers to account. The key role
assigned to communities in targeting relief is also in line with the principles of the DRM policy on
decentralisation and community empowerment.

DRM Principle 3: Decentralized and community-based

(see Table 3 in Part Two)

A ‘community’ is defined in the DRM Policy as “a group of people with a commonality of association,
having common interest, shared experience, or function and living in the same locality and under the
same public administration.” For targeting purposes, a community must be small and coherent enough
that its members and representatives have a good knowledge of each other’s livelihoods and living
conditions, and can reasonably be expected to make informed judgements about who is most in need of
assistance. In sedentary farming areas this unit is taken to be the village (e.g. the ‘got’ in Amhara, ‘ganda’
in Oromiya, ‘gasha’ in SNNPR or ‘kushet’ in Tigray). In pastoralist societies, the appropriate community
unit may be area-based or it may be a clan or other social structure. Regions should identify the most
suitable community unit for the targeting process set out below, according to the culture and social
organisation of their populations.
‘Pure’ community targeting, in which the community has complete decision-making autonomy over how
and whether to target relief assistance, is not considered suitable for the implementation of national
policy on the targeting of relief food assistance. In order to ensure an acceptable level of consistency,
adherence to national policy, accountability and awareness-raising about the correct use of emergency
relief assistance, administrative guidance and oversight of the targeting process is required.

51
TARGETING RELIEF FOOD ASSISTANCE PART FIVE: BENEFICIARY TARGETING

On the other hand, ‘pure’ administrative targeting in which decisions are made without the participation
of beneficiary communities is also considered unsuitable, partly because purely administrative (objective,
‘outsider’) targeting methods require a high level of capacity and resources, as explained above and in
Part Three. More importantly, it undermines the principle of community empowerment and
decentralisation.
Therefore, a combination of administrative and community targeting elements will continue to be the
basis of relief targeting at beneficiary level. It is recognised, however, that the implementation of this
approach has not been without problems in the past, and its effectiveness has varied greatly from place
to place. Continued efforts are needed to improve the system. These guidelines aim to strengthen:
 the empowerment of communities through more consultation, better information flows, and respect
for community values in tailoring the targeting criteria and processes to the local context;
 the participation and contribution of humanitarian partner organisations in targeting relief from
wereda to beneficiary level; and
 accountability and procedures for the correction of errors or abuses.

19.1. Situations where community targeting may not be applicable

Combined community-administrative targeting is the preferred method of household targeting in the
Ethiopian relief system. However, there are some circumstances in which community targeting may not
be appropriate, in which case the procedures set out here may be replaced by other targeting methods
(or a short-term blanket distribution if the emergency warrants it). A judgement should be made about
the most appropriate, effective and efficient targeting approach for each emergency, depending on the
situation and the nature of people’s needs.
 The major exception to the community-administrative targeting process is in the distribution of
supplementary foods (SF), either to vulnerable group members or to malnourished individuals (see
Section 24 below). In these cases, targeting criteria are not decided by the community but are fixed
and standardised by international nutritional practice and national technical guidelines. In blanket SF
where nutritional screening is not used but rations are provided for nutritionally vulnerable groups
(children under five, pregnant and lactating women, and the elderly), the community may be asked
to help in registering the members of these groups, but they do not set the criteria.
 In situations of conflict or severe disruption to normal life, community institutions may not function
well or may not represent all those in need of assistance.
 In some fast-onset disasters, community-administrative targeting (or indeed, any targeting at the
household level) may be inappropriate or unnecessary, or it may be too slow.
In these situations, the full community targeting process may be suspended. However, the beneficiary
community should always be informed and if possible consulted about any planned relief distributions,
and about the targeting or allocation rules (including ration entitlements). The principles of transparency
and accountability still apply.

19.2. Roles of government, partners and community

Table 12 summarises the responsibilities of the administrative and community elements in the targeting
process, and the range of support roles which can be played by humanitarian partners (depending on
their level of field presence and their operating agreements with the relevant authorities). The DRM
Policy supports the active participation of humanitarian partner organisations in emergency response
activities.

DRM Principle 4: A participatory approach

(see Table 3 in Part Two)

52
TARGETING RELIEF FOOD ASSISTANCE PART FIVE: BENEFICIARY TARGETING

Table 12: Administrative, partner and community roles in relief targeting

ADMINISTRATION HUMANITARIAN PARTNERS * COMMUNITY

Assess and prioritise emergency Participate in assessments and Provide information and feedback to
relief needs. committees. EW and needs assessment
Mobilise resources. Share relevant information arising systems.
Allocate quantities of relief and from field operations (e.g. nutrition
number of people who can be surveys, food security information,
assisted by wereda/ kebele / community reports).
community.
Facilitate community discussions, Decide details of registration
two-way information flows and procedure and community
awareness-raising about the correct Help to facilitate community representation.
use of relief assistance. discussions, awareness-raising, Discuss and give feedback on
Provide information about planned and information flows, jointly with needs and relief assistance.
distributions (timing, quantity, etc.). administration or as delegated by
wereda relief committee.
Define broad targeting criteria (as Fine-tune / decide detailed targeting
set out in these guidelines). Contribute staff time and resources criteria, according to local
to support administrative oversight livelihoods, values and social
of community targeting. conditions.
Strengthen reporting and record-
Oversee the community targeting Identify and register eligible
keeping of community targeting
process (to ensure fairness, beneficiaries.
process.
transparency and accountability). Verify beneficiary list through public
posting / meeting.

Maintain and compile registration Assist with registration and record- Provide a copy of community
lists and records of the targeting keeping, e.g. through training, registration lists to the kebele /
process, including any appeals and knowledge transfer, capacity wereda committee and keep a copy
action taken. Regularly compile development, or support for at community level for consultation
these records and forward them to recurrent costs. or checking as needed.
Regional authorities. Share records
with partners, and provide them to
monitoring / audit teams on request.
Monitor targeting process and Participate in monitoring and Correct errors at community level if
outcomes (through regular appeals system. possible.
monitoring, spot-checks or audit). Share information from regular Report any problems or
Ensure accountability, investigate monitoring and evaluation. malpractice.
complaints and correct problems. As members of wereda relief bodies
or assessment missions, conduct
spot checks and audits on targeting
process and appeals.
Feed back any problems observed
to authorities for action.

* Humanitarian partners are UN, donor, non-governmental and civil society organisations participating with the
government in providing relief food assistance. Their potential to support the targeting process from wereda to
beneficiary level varies widely, depending on their field presence and their operational mandates. Therefore, this
column contains suggested areas where they may be able to make a valuable contribution of expertise and/or
resources. Details should be agreed between partners and the relevant authorities for each relief operation.

53
TARGETING RELIEF FOOD ASSISTANCE PART FIVE: BENEFICIARY TARGETING

20. Process and governance of community-administrative targeting

20.1. Institutional arrangements

20.1.a. General lines of communication and accountability

Each Region is requested to decide and establish the detailed structures and two-way lines of
accountability and information flow which are necessary to carry out the roles and responsibilities set out
in these guidelines, from the Region to the community level. The general pattern of these arrangements
is represented in Figure 14. The details should be decided at Regional rather than Federal level for three
reasons:
 the principle of increasing decentralisation of emergency relief to the Regions;
 the variation in structures and names of the relevant government bodies in each Region since the
BPR (Business Process Re-engineering); and
 the variation in humanitarian partners’ operating arrangements and field presence in different parts
of the country.
It is important for accountability that responsibility for each step in relief targeting is assigned to a
specific office or individual, and that the participatory principle of including humanitarian partners in
relief targeting is clearly communicated to the weredas. Once the Regional details of the institutional
chain are agreed, they should be set out in writing and shared with the federal DRMFSS, all disaster-
affected weredas in the Region, and relevant humanitarian partners.

Figure 14: General lines of accountability and information flow from Region to community

REGION PARTNERS

WEREDA PARTNERS

Disaster-affected
COMMUNITY

20.1.b. Wereda and kebele responsibilities

The wereda (district) and kebele (sub-district) are the levels of local government closest to beneficiary
communities, and therefore have key roles to play on the administrative side of community-
administrative targeting as set out in Table 12. Emergency relief assistance is by its nature temporary and
unpredictable: therefore, it is not necessary to establish permanent local government structures to
manage it. When a disaster response operation is planned in a given wereda, a temporary ad-hoc body 25
should be formed first at wereda level to take responsibility for the targeting and other aspects of the
planning and distribution, for the duration of the operation. This ad-hoc relief body could be called a
committee, a task force, a team, or any other term consistent with the organisation of wereda functions.
It may be an existing committee sitting in special session, or a delegated sub-committee of an existing
body, or a new grouping of key people convened for a specific emergency. Each Region should review

25 „Ad hoc‟ means for a specific purpose.

54
TARGETING RELIEF FOOD ASSISTANCE PART FIVE: BENEFICIARY TARGETING

and decide the appropriate form and name of these temporary wereda relief bodies as part of their
overall institutional review of responsibility and accountability for relief targeting.
A major responsibility of the wereda is the allocation of relief resources to kebeles. This should always be
based on an objective assessment of relief needs in each kebele (see Section 17.2. on Level 2
geographical targeting). Relief should not be spread equally or by ‘quota’ among all kebeles in the
wereda, unless there is clear evidence that all kebeles are equally in need. Information from the early
warning system and needs assessments should be taken into account. The HEA system can potentially
help in prioritising and quantifying the needs of different kebeles by livelihood zone within a wereda (see
Section 18.2). Detailed guidelines for weredas on this level of targeting should be included in Regional
implementation handbooks.
When a kebele is included in the relief distribution plan, an ad-hoc relief targeting body (committee, task
force or team) should also be formed at the kebele level. At both wereda and kebele level, humanitarian
partner staff should be included in the relief body if they are available, to ensure smooth co-ordination
and information exchange. The suggested members of these bodies are indicated in Table 13.
In PSNP weredas, it is recommended that the existing wereda, kebele and community Food Security Task
Forces (FSTFs) be mandated to manage relief targeting, in order to make best use of existing capacity and
to ensure smooth coordination between emergency relief and the PSNP’s Contingency and Risk Financing
mechanisms (see Section 25 below). Additional members, including staff of humanitarian partner
organisations, may be co-opted to join the FSTFs for the duration of the relief operation as needed.
If possible, the gender balance of the relief bodies should be equal between women and men. However,
this may be an unrealistic goal in most parts of the country at present. At wereda level, there are few
women in senior positions and therefore few women will be included in relief bodies ex officio (that is, on
the basis of their official position). A more equal gender balance may be achieved at kebele level,
because there are more women among the government change agents (DAs and Health Extension
Workers) than among wereda departmental heads.

Table 13: Recommended membership of wereda and kebele bodies responsible for relief targeting

Wereda Kebele

 Representative(s) of the wereda council  Representative(s) of the kebele council

 Department Heads responsible for:  Development Agents (DAs)
o Food Security  Head of Women’s and Youth Associations
o Early Warning  Health extension workers or volunteers
o Women’s Affairs
 Director of primary school or other respected
o Health
local figures
 Representative(s) of all humanitarian partner
 Staff of partner organisations (if available at
organisations involved in relief within the
kebele level)
wereda
 Elected community representatives
 Others as decided by the council
 Others as decided by the council

The most important thing is to ensure that women’s concerns are fairly represented, and that their needs
and vulnerabilities are adequately considered, in the targeting of relief assistance. If equal
representation of women and men is not achievable, each wereda and kebele relief body should include
at least one woman member, i.e. the head of the Women’s Affairs Department (at wereda level) and the
head of the Women’s Association (at kebele level). It is recommended that these representatives be
given a special mandate and responsibility to represent and promote the interests of women in the relief
programme, and to receive complaints or appeals from women community members. Each Region

55
TARGETING RELIEF FOOD ASSISTANCE PART FIVE: BENEFICIARY TARGETING

should develop detailed plans for this in consultation with the Regional Women’s Affairs Bureau (or
equivalent).
The functions and responsibilities of the wereda and kebele relief bodies should be set out in more detail
in the implementation handbooks, in line with the general administrative roles summarised in Table 12.
These roles can be broadly categorised as needs assessment; oversight and facilitation of community
targeting; ensuring information flows, awareness-raising and accountability; monitoring and correction of
problems; reporting and record-keeping.
To be carried out effectively, all of these roles require resources, including administrative budgets for staff
time, transport, communications, and consumable supplies such as reporting and registration formats;
longer-term institutional capacity development and skills transfer; and recurrent ‘top-up’ training at the
beginning of each relief operation to refresh knowledge and to counteract the high rate of staff turnover
in local government.

20.1.c. Community representation

The preferred structure at community level is an elected relief committee representing all sections of the
community, combined with open meetings at which all community members can exchange information
with government and partner representatives, and can discuss and challenge the proposed targeting
criteria and registration lists. Equal numbers of men and women should be elected, and care should be
taken to represent all sections of the community (including any vulnerable or marginalised groups). All
elected representatives, especially women, should be consulted about the time and place of committee
and community meetings. Sometimes women are elected but are unable to attend because meetings
conflict with their domestic work: in this case the election of women is a mere token and does not ensure
representation of women’s concerns. A procedure should be in place for the community to replace any
elected representative who does not carry out their responsibilities honestly, fairly and diligently.
This is broadly the model of community representation which has been taken up and strengthened by the
PSNP, and the emergency relief system should follow this model wherever possible. In PSNP weredas, it
is recommended that existing Community Food Security Task Forces be given responsibility for relief
targeting (see Sections 20.1.c. and 25). The election of community committees for relief targeting should
be supported and monitored by members of the wereda or kebele committees (see above), including
humanitarian partners if possible, to ensure that the principles above are understood and adhered to.
Members of the wereda or kebele committees should also be present at community meetings whenever
possible, to exchange information and to be available to hear any complaints about the targeting process.
This model should be applied flexibly and may be adapted to local situations. The most effective
arrangement for community representation may depend on existing social structures and norms for
collective decision-making, and on the degree of local experience with community-based relief or
development. For example, in some communities there is no tradition of speaking openly in a full
community meeting, especially for less powerful community members. In others, respected elders or
established leaders may be more effective in carrying out the targeting process than elected
representatives. In contexts where it is not acceptable for women and men to sit together on a
committee, or where women may not feel able to speak or to be heard in a joint committee, it can be
more effective to have separate women’s and men’s committees. Each can then discuss freely among
themselves, and their views can be transmitted to the community or to a higher-level committee by a
delegated representative.
Each Region should therefore review best practice and successful examples of community decision-
making among their own populations, in consultation with the weredas and with experienced
humanitarian and development partners, and develop detailed local guidelines on community
representation in the targeting process.

56
TARGETING RELIEF FOOD ASSISTANCE PART FIVE: BENEFICIARY TARGETING

Box 4: Seven key steps in the community-administrative targeting process

1. CONSULTATION AND 2-WAY INFORMATION-SHARING ON DISASTER IMPACTS AND NEEDS

Representatives of the wereda / kebele relief committee should begin by explaining to the community the needs assessment
findings for their area, and encourage two-way information sharing by eliciting community views on:
 Disaster events currently affecting this community
 Which food and income sources are affected, how, and to what extent?
 Which groups of people rely on these sources?
 What coping strategies (i.e. alternative / additional / expanded sources of food and income) are available to people in
this community? Who has access to them?
 Which of these coping strategies are acceptable and sustainable, and which are considered unacceptable or damaging
to people‟s health, dignity or future livelihoods?
 Considering all these factors, what groups or kinds of people are worst affected by the current disaster in this
community?

2. EXPLANATION OF THE PURPOSE OF RELIEF FOOD ASSISTANCE

Explain that relief food assistance is provided:
 for a limited period, because of the specific disaster temporarily affecting this area, and
 in limited quantities, for people who cannot meet their family‟s short-term basic food needs. It is not an entitlement for
everyone in a disaster-affected area, or for everyone who is not in the PSNP. People who can manage without relief
are requested to exclude themselves from the targeting.

3. INFORMATION-SHARING ABOUT RESOURCES ALLOCATED TO THE COMMUNITY

Provide the following information, and display it in a public place accessible to all community members:
 number of people who can be registered in this community (i.e. quantity of relief / number of rations allocated)
 ration per person per month or per distribution round (display posters)
 number and timing of planned distribution rounds (including start and end date of relief operation)
 explain principle of registering all household members
 explain principle of registering women as household food managers
 if SF is planned, explain purpose and targeting rules

4. DISCUSS AND AGREE LOCAL TARGETING CRITERIA

Based on all the above information and discussion, facilitate community discussion about appropriate local criteria for
registering beneficiaries:
 explain basic (national) criteria for relief food assistance
 discuss provisional criteria suggested for this area by HEA / needs analysis, i.e.
o livelihood profiles (food and income sources most affected by the specific hazards)
o wealth group characteristics (asset levels etc. of people likely to be worst affected)
o coping strategies or resources that would enable people to meet the food deficit themselves
 discuss which coping strategies are locally acceptable / sustainable, and which are damaging
 discuss vulnerable groups – who is least able to cope for themselves?
o e.g. elderly, disabled, female household heads, orphans etc. – respect community values, who do they see
as most vulnerable? How are these people usually supported in this community?
 based on all the above, agree inclusion and exclusion criteria for relief assistance in this community
 agree process for identifying and registering eligible people

5. COMMUNITY PRODUCES DRAFT REGISTRATION LIST (according to process agreed)

6. COMMUNITY DISCUSSION, ADJUSTMENT AND ENDORESEMENT OF LIST

 Draft registration list should be made accessible to all community members – ideally it should be read out in an open
meeting and posted in writing for an agreed period during which appeals and changes can be made.
 Community members should have time and opportunity to raise objections and disagreements – in an open meeting, or
by application to a designated appeals committee, or to staff of local government or humanitarian partners engaged in
monitoring and audit of the targeting and distribution.

7. FINAL VERSION OF REGISTRATION LIST SUBMITTED TO KEBELE & WEREDA RELIEF COMMITTEE

57
TARGETING RELIEF FOOD ASSISTANCE PART FIVE: BENEFICIARY TARGETING

20.2. Key steps at community level

Box 4 summarizes the key steps recommended for the process of beneficiary targeting at community
level. This is intended as a checklist for use by representatives of the wereda and kebele committees
(from both government and partner organisations) who are responsible for facilitating the community
process. It is not suggested that all these steps should be implemented in one meeting. The details and
the order of discussions, as well as the institutional format of community representation (see above) may
be adapted or added to, according to the local situation.
Step 4 (agreeing local criteria) is treated in more detail in Section 21 below. The process for Step 5
(identifying and registering eligible households) can be managed in a number of ways. Options include:
a) At a full community meeting, people propose themselves or others for inclusion in the list, and then
the whole community discusses and decides on each case.
b) The list is drafted by the community committee (or other agreed representative body), and then
presented to a full community meeting for discussion and adjustment. The draft list may be based
on prior knowledge, or on house-to-house assessments according to the agreed criteria.
c) Households are ranked according to relative need, with or without triangulation, by the community’s
representatives or by groups representing different sections of the community. Ranking has the
advantage that it does not need to be repeated if the ‘quota’ of relief changes: the cut-off point can
simply be moved up or down the ranked list.
If possible the community should choose which process to use, or if necessary the administrative
representatives can suggest or prescribe how it should be done. In either case, all community members
should be informed of the process and the criteria.

SPHERE Key Indicator 2:

“Targeting mechanisms are agreed among the affected population (including representatives
of vulnerable groups) and other appropriate actors. Targeting criteria are clearly defined and
widely disseminated.”
(see Box 3 in Part Two)

20.3. Transparency: information is power

Principle 7 of the DRM Policy states that information is power. Like other forms of power, information
only works when it is flowing. Transparency at community level means that information about local food
security conditions, the quantity and timing of relief assistance provided and the targeting process should
flow in two directions (from and to the community). Information from the community, and dialogue
about the needs assessment and targeting processes, is highly valuable in strengthening the early
warning system, providing feedback on the effectiveness of the emergency response system, and
fostering ‘joined-up targeting’.
Genuine community participation in the targeting of relief requires that communities should understand
the purpose and correct use of emergency relief, and be informed of their rights and responsibilities with
regards to its targeting and distribution. All community members should understand the process and
criteria used in targeting, and should know who they can complain to if they believe they have been
unfairly excluded or have any other complaints about the targeting process. Beneficiary lists should be
publicly posted and accessible to all community members (see Box 4 above). All beneficiaries, as well as
community representatives, should know the quantity and timing of assistance that has been allocated to

58
TARGETING RELIEF FOOD ASSISTANCE PART FIVE: BENEFICIARY TARGETING

them. Good information and awareness-raising campaigns can greatly increase the effectiveness of
targeting. Information can also empower marginalised target groups to claim assistance.
For all these types of information to be available to the community, they must of course first be available
to the local government and partner organisations engaged in the community-administrative targeting
process at wereda level. Regions should ensure the timely exchange of such information among all
relevant stakeholders as far as the wereda level. The wereda council and relief committee are
responsible for information flows within the wereda.
Accountability (see below) requires that information from community members about any problems with
the targeting system has channels to flow through, and will be heard and recorded. Only then can
complaints be investigated and acted on: transparency is a necessary condition for accountability.

DRM Principle 7: Information is power

(see Table 3 in Part Two)

20.4. Accountability
Accountability means that decisions can be checked and challenged, and that decision-makers can be
held responsible for their actions. Upwards and downwards accountability is needed: that is, all decision-
makers and organisations involved in targeting are ultimately responsible to the people in need of
assistance, as well as to the institutional hierarchies above them.
Accountability requires:
 an independent appeal channel (that is, independent of the committees and representatives who
make the targeting decisions) for community members who believe they have been wrongly
excluded, and for any complaints about the fairness of the process;
 procedures for removing decision-makers who do not carry out their responsibilities correctly,
honestly and diligently;
 a free flow of information about the targeting process and criteria, and about the timing and quantity
of distributions, among local government, partners and community; and
 proper record-keeping of registration lists, targeting processes, and appeals.

DRM Principle 5: Accountability and Responsibility

(see Table 3 in Part Two)

20.4.a. Correcting errors: appeals and complaints

Key step 6 in the community targeting process (see Box 4 above) enables community members to
challenge and revise targeting decisions before the registration list is finalised and forwarded to the
kebele and wereda. If disagreements remain after this process, or if community members have any
complaints about the targeting process or the actions of any decision-makers, these complaints should be
heard, recorded, investigated and (if appropriate) acted upon.
Appeals bodies should be established for this purpose at kebele level, as set out in the previous Food Aid
Targeting Handbook (DPPC 2002), for the period of the relief operation. Members of the kebele appeals
body should be respected elders, leaders (e.g. religious leaders), professionals (e.g. teachers or health

59
TARGETING RELIEF FOOD ASSISTANCE PART FIVE: BENEFICIARY TARGETING

workers), and elected representatives of communities receiving relief assistance. Equal numbers of
women and men should be aimed for. If this is not achievable, at least one woman should be included in
the appeals body, and she should have special responsibility for receiving and following up appeals from
women. The kebele appeals body should consider and keep records of all appeals and complaints, and
make recommendations for action to the relief committee when appropriate. Anybody who feels their
appeal or complaint has not been satisfactorily dealt with by the kebele appeals body should then take it
to the wereda council. Community members should not be expected to travel further than the wereda
capital to make an appeal or complaint.
In PSNP weredas where Kebele Appeals Committees (KACs – see PIM section 7.2.2.) are already
established and working effectively, it is recommended that they should also be given responsibility for
appeals relating to relief assistance. Additional members may be co-opted, or a sub-committee formed
to deal specifically with relief, as needed.
Ideally, appeals bodies should be independent of the original decision-makers, and therefore there
should be no overlap in membership with the kebele relief committee. However, this can be difficult to
achieve, since the same few influential people or groups can dominate both the decision-making and the
appeals process. In implementation, the effectiveness of these appeals bodies so far has been very
mixed.

Figure 15: Lines of appeal: checks and balances

Decision-makers Regular appeal bodies Spot-checks / monitoring / audit

REGIONAL
Authorities

WEREDA WEREDA
Relief body Council / Cabinet

KEBELE KEBELE
Authorised staff from
Relief body Appeal Body
government and
humanitarian partner
organisations
COMMUNITY
(Relief Committee
or full assembly)

a b c

Community members (beneficiaries and non-beneficiaries)

a. Appeals and complaints should first be referred back to the decision-making body.
b. If not resolved, or if appellants are unwilling to go to the decision-makers, complaints should be taken to
the kebele appeal body or the wereda council.
c. At any time, community members can refer appeals or complaints to authorised staff engaged in food
distributions, monitoring or assessment missions.
solid line = recommendations and action

60
TARGETING RELIEF FOOD ASSISTANCE PART FIVE: BENEFICIARY TARGETING

To strengthen the checks and balances in the appeals process, staff of government agencies and
humanitarian partners engaged in the disaster response system should be authorised to receive appeals
or complaints whenever they are present in a targeted wereda. Each Region should develop a detailed
reporting format for such cases and an agreed procedure for them to be forwarded to the wereda, and if
necessary to a designated responsible person at Regional level, for follow-up. Action taken in each case
should be recorded. Such authorised staff in the field should also be given a mandate to make spot-
checks and audits26 of the targeting process, in addition to the routine monitoring system. If requested
they may also support and advise members of the kebele appeals body and the community relief
committees, who have extremely difficult tasks and can find themselves pressured from many sides.

20.4.b. Replacing bad decision-makers

In any situation where valuable resources are being allocated, there will inevitably be risks and
opportunities for abuse or poor management. It is not possible to eliminate all such risks. The important
thing is to have systems in place to detect any problems and to correct them. Therefore, accountability
requires that poor decision-makers can be replaced.

DRM Principle 13: Avoid malpractice

(see Table 3 in Part Two)

At community level, a procedure should be in place for any member(s) of the elected relief committee to
be replaced if their constituents are not happy with the way they are carrying out their responsibilities.
Any complaints about the community relief committee should go first to the kebele appeals body or
authorised staff, and then to the wereda who should investigate and organise a re-election if necessary.
Complaints against any officials or employees should be dealt with through the same channels if possible,
and then through legal procedures to be defined by Federal government in support of the DRM Policy.

20.4.c. Record-keeping, registration and ration cards

The accountability of the relief targeting system has been seriously hampered in the past by poor record-
keeping, inaccurate or unavailable registration lists, and in most cases the absence of ration cards which
provide both beneficiaries and monitoring staff with a clear record of relief assistance allocated and
received. Practices such as rotation of beneficiary lists, under-registration of household numbers, or
splitting of households to qualify for registration are unacceptable and should be corrected. In addition
to distorting the targeting itself, they make registration lists almost impossible to interpret as an accurate
record of who received what.
In PSNP weredas, it is recommended that a ‘single registry’ system be introduced, whereby a single list or
data-base will record all transfers received by each registered household, whether from regular PSNP, CB,
RFM or relief.
While considerable effort has been spent on recording and accounting for quantities of commodities at
every stage of distribution, relatively little has been invested in tracking who has received the assistance.
Every effort should be made to improve these aspects of the targeting system. Resources should be
sought from donors and partners for this purpose if necessary.

26 „Audit‟ means examination and verification of records.

61
TARGETING RELIEF FOOD ASSISTANCE PART FIVE: BENEFICIARY TARGETING

21. Fine-tuning the local targeting criteria

Part Two (Core Principles) defined the target group for relief food assistance in principle as:

People whose access to food (through purchase, production or any other means) is
temporarily reduced by a shock or a slow-onset disaster,
and they are unable to maintain an adequate nutritional intake,
or they are able to maintain an adequate nutritional intake only by resorting to
unacceptable or damaging coping strategies, risking irreversible damage to their health
or livelihoods.

Part Three (Toolbox) outlined the types of criteria that can be used in targeting relief food assistance –
socio-economic, nutritional, or categorical (group characteristics), direct measurement or proxy
indicators. It also noted that a combination of several criteria, and a balance of inclusion and exclusion
criteria, is likely to produce fairer and more accurate targeting outcomes than a single indicator.
In Part Four it was explained how the different levels of the relief targeting system and the HEA analytical
process help to set provisional criteria which describe the households or groups who are expected to be
most affected by a given disaster, and who should therefore be prioritised for relief assistance.
At the final (beneficiary) level of the system, the community, through the process described above, has to
interpret these broad principles and provisional characteristics in terms of their own livelihoods and
social situation, to decide the detailed criteria which will correctly identify the people most in need of
relief in practice.

21.1. Targeting according to need

The fundamental principles of targeting emergency relief are that resources should be allocated on the
basis of need alone, and that priority should always be given to the most vulnerable people and those in
most urgent need. At the community level, this means that the needs and coping capacity of all groups in
the community should be considered. No factor other than need (such as family ties, friendship, enmity
or prejudice) should be allowed to influence the selection process. Relief should not be shared equally
among everyone, but should be allocated equitably (fairly) according to people’s different levels of need.

DRM Principle 9: Impartiality

(see Table 3 in Part Two)

Following the logical framework of the Household Economy Approach, a household’s food access gap will
depend on a combination of:
 the impact of a specific disaster event on the food and income sources the household usually
depends on;
 the wealth or assets the household has to fall back on; and
 the coping strategies and social support networks it can draw on to fill the food gap.
The community discussion process recommended in Steps 1 and 3 of Box 4 above lays the groundwork
for deciding detailed locally-appropriate criteria for each of these aspects. Table 14 suggests some
factors to consider in setting inclusion and exclusion criteria to match the broad criteria set by these
guidelines. Community representatives should be encouraged to adapt or add to these suggestions
according to their local knowledge, and to use a combination of different criteria rather than a single
indicator such as an asset threshold. At the end of the process if they are in doubt about a household’s

62
TARGETING RELIEF FOOD ASSISTANCE PART FIVE: BENEFICIARY TARGETING

eligibility, they should make a judgement on whether the household meets the broad definition. If the
community finds it possible and acceptable, the number of distribution rounds for each household can be
adjusted according to their level of need (for example, some might need only 2 months’ relief while
others need 6). This should be clearly recorded on the registration list. However, all members of eligible
households should be registered, and beneficiary lists should not be rotated.

Table 14: Broad and detailed criteria for household targeting

Detailed criteria
Broad criteria to be decided by community, with guidance and facilitation from wereda and
kebele relief committees
(National Guidelines)
Factors to consider:

inclusion criteria exclusion criteria

Households whose food and Livelihood groups or wealth groups Other food and income sources not
income sources are affected by the relying on badly affected food and affected by the disaster (including
disaster… income sources transfers from on-going
development or safety net
programmes)

… and cannot meet their basic food Household size, labour, and Sufficient stocks or savings
needs over the coming months or dependency ratios (food needs
season compared to earning capacity)
Vulnerable groups (e.g. widows,
orphans, chronically ill or disabled)
Months of food gap

… or will have to rely on damaging Very few assets*; selling them would Access to acceptable coping
or unacceptable strategies to meet make recovery difficult strategies
their food needs
Forced to rely on distress strategies Enough assets to sell some without
which are dangerous, unacceptable risking destitution
or damaging
Support from family or social
networks (e.g. remittances)

* The type and number of assets should be locally defined; see HEA baselines and wereda posters for provisional criteria

In deciding which coping strategies are damaging or unacceptable, environmentally damaging strategies
should be discouraged. Wereda agriculture and natural resource experts should advise on this.

DRM Principle 2: Protection of natural resources

(see Table 3 in Part Two)

21.2. Respecting community values

Every community has its own values and institutional norms for supporting the poor and needy, and for
mutual aid in hard times. These values may be based on religious practice, or on cultural and social
custom. As far as possible relief assistance should avoid weakening existing institutions of informal social
protection and redistribution, and criteria for targeting should take account of local values and

63
TARGETING RELIEF FOOD ASSISTANCE PART FIVE: BENEFICIARY TARGETING

sensitivities. Consulting the community about their own perceptions of who is vulnerable and in need of
help, and how such people are usually supported in their community (Step 4 in Box 5 above), opens the
discussion of targeting criteria to include and respect cultural norms. In some cases existing institutions
and traditional means of ensuring aid to those who need it may be used to channel relief assistance,
provided that the outcome is consistent with the broad criteria in Table 14 and the general principles in
these guidelines. If community values clash with the principles of targeting according to need and
prioritising the most vulnerable, then the principles take precedence.

DRM Principle 10: Respect cultural norms

(see Table 3 in Part Two)

SPHERE Key Indicator 3:

“Targeting mechanisms and criteria should not undermine the dignity and security of
individuals, or increase their vulnerability to exploitation.”
(see Box 3 in Part Two)

21.3. Avoiding dependency traps

Targeting criteria should not create ‘dependency traps’, which give people a rational incentive to
compromise their future livelihood and food security in order to qualify for short-term assistance. The
best-known example of this in Ethiopia is the previous practice (in ox-plough farming communities) of
disqualifying households from relief if they owned one ox. This was intended as a simple wealth-test to
identify the poorest, but there are many anecdotal reports of people selling their last ox to qualify for
food aid, thus reducing their ability to plough for the next season and sacrificing their future livelihood to
achieve short-term survival.
This type of trap can push people into destitution and chronic food insecurity, damaging their future
quality of life and, in aggregate, compromising the growth and development aspirations of the country as
a whole. This is contrary to the objectives of the DRM and development policies. Relief food assistance is
intended to help people get through a period of acute food insecurity, without irreversible damage to
their health or their livelihoods.
In setting local targeting criteria, community representatives (with guidance from local government and
partners) should therefore avoid creating such traps by:
 not relying on a single indicator or a single resource type (such as livestock) to select
beneficiaries;
 considering all the household’s coping capacities and resources (including labour), and assessing
their overall situation as explained above;
 providing relief to people before they have been forced to employ damaging strategies such as
selling irreplaceable productive assets.

DRM Principle 11: Minimize dependency and promote livelihood recovery

(see Table 3 in Part Two)

64
TARGETING RELIEF FOOD ASSISTANCE PART FIVE: BENEFICIARY TARGETING

22. Registering women as food entitlement holders

In targeting relief food assistance, the international standard practice applied by WFP and other agencies
is to register women as the named beneficiaries or ‘food entitlement holders’ for relief distributions,
whether or not they are household heads. The targeting unit is defined as the named woman plus the
children and adults she usually cooks or provides food for. This approach has a number of advantages
over the present practice of registering household heads:
1. It recognises and respects the important role of women as household food managers. It is women
who store and ration food within the household, and eke out supplies in times of shortage by the
addition of wild foods or other strategies. Women are therefore in the best position to manage relief
food supplies for the benefit of household members.
2. It can help to empower women and raise their status. However, it is not intended to challenge the
authority of men or to cause conflict. Women beneficiaries in male-headed households are
registered as household food managers, not as household heads.
3. International evidence shows that food delivered directly to women is more likely to reach the
children and other vulnerable members of the household.
4. It is applicable in almost all social contexts, and avoids some types of confusion or unfairness that can
arise when registering household heads – for example, in polygamous communities (where each wife
should be registered as a separate food entitlement holder with the people she cooks for, and the
husband should be registered as a member of one of these units), or in situations where household
heads may be absent (such as displaced camps, or pastoralist communities when the men have
migrated with the herds).
5. It may help to differentiate relief from other types of transfer or development project, and to raise
awareness that the primary purpose of relief food assistance is to ensure adequate short-term food
access for disaster-affected people.
The registration of women as food entitlement holders should be introduced in a sensitive and flexible
way, under the guidance of each Region and in consultation with partner organisations and communities.
Consultation with women beneficiaries themselves is especially important, to ensure that new
procedures do not generate unnecessary conflict or in any way endanger women or add to their
workloads. For example, registration in women’s names does not necessarily mean that they must
collect and carry the food themselves. There is a large body of international experience with this
approach which can be drawn on for best practices and ways to avoid known pitfalls (see, for example,
WFP’s Guidelines on Implementing the Enhanced Commitments to Women).
The recommendation to target women in this way has been in the government’s Gender Mainstreaming
Guidelines since 2004, but it has not been implemented. Ethiopia should now align its household
targeting with international practice and begin to implement relief registration in women’s names.

 See FDPPC 2004, Gender Mainstreaming Guidelines for Disaster Management Activities

23. General food distribution (GFD)

In-kind relief food assistance is of two main kinds: General Food Distributions (GFD), which are intended
to cover basic nutritional needs for all targeted beneficiaries, and Supplementary Foods (SF) which are
additional rations intended for groups with higher nutritional needs (especially children under five years

65
TARGETING RELIEF FOOD ASSISTANCE PART FIVE: BENEFICIARY TARGETING

old, and women who are pregnant or breastfeeding).27 Therapeutic Feeding is a specialised type of
supplementary nutrition provided under medical supervision to treat severe acute malnutrition (SAM). It
is not included in these guidelines because it should always be targeted by health professionals, and falls
under the responsibility of the Ministry of Health.

23.1. Ration rates and the ‘food basket’

The standard GFD ration is based on an average daily energy requirement of 2,100 kcal per person. This
is a global average, set by the World Health Organisation, and is normally used as an initial planning
figure in emergency situations unless good information is available about the actual nutritional needs of
the target population (in terms of their age and gender composition, activity levels, health status,
environmental conditions, etc.) and their access to other food sources apart from relief.

Box 5: Standard full food basket for GFD in Ethiopia

15 kg cereals
0.45 kg vegetable oil per person per month
1.5 kg pulses

The standard full ‘food-basket’ for Ethiopia is shown in Box 5. These amounts are set on the basis that “it
is expected that the population in need of assistance will be dependent on relief assistance” (DRMFSS
2010b:12). If resources are sufficient, it is recommended that the full basket is provided for all members
of targeted households: this ensures an adequate nutritional intake, and in many situations will provide
some degree of livelihood protection.28 It also aids transparency and accountability if there is a single
standardised ration, because it is easier to ensure that the community and beneficiaries understand the
amounts allocated and can therefore challenge the authorities if the amounts they receive do not match.
However, in reality it is often the case that the commodities for the full basket are not available, or
resources are insufficient to provide a full ration to everyone who has been registered for relief. In this
case, decision-makers have two basic options: to cut the number of beneficiaries, or to cut the ration per
person. In making this judgement it is important to consider the beneficiaries’ access to other food
sources and the estimated size of actual household food deficits.
For slow-onset food crises among populations who are not displaced and are still pursuing their
livelihoods – the most common type of relief context in Ethiopia –people are rarely entirely dependent
on relief assistance unless the aid comes very late and the emergency is at an advanced stage. As
explained in Part Four, the HEA method of needs analysis quantifies the impact of shocks on the food and
income sources of typical households in different wealth groups, resulting in an estimated percentage
shortfall in household food access. Very rarely is this gap 100% of the household’s food needs.
Therefore, less than a full ration may be enough to meet their food gap (provided the targeting is
sufficiently accurate and timely). More use could be made of the HEA data in making decisions about the
quantities of relief needed per targeted household: however, this should also be triangulated by
discussion with the community.
It should also be kept in mind that the total amount of aid actually received by beneficiaries, and
therefore the effectiveness of targeting in meeting their food deficit, is a product of the ration and the

27 These groups are targeted for supplementary feeding because, during these key periods in their lives, they need a
higher nutritional intake than the general population average. The elderly are also often included in
supplementary feeding because they are highly vulnerable to malnutrition and less able to cope for themselves in
emergencies.
28 That is, it will protect beneficiary households from needing to sell productive assets or cut spending in key areas
such as education and health. They will therefore be able to recover more quickly after the disaster period, and
will have greater resilience to future shocks.

66
TARGETING RELIEF FOOD ASSISTANCE PART FIVE: BENEFICIARY TARGETING

number of distribution rounds received. An adequate benefit level depends on the quantity received
over the whole period of the emergency, compared to the household food gap (see Section 14.3. in Part
Three for an explanation of benefit-level errors).
In cases where the beneficiary population is temporarily entirely dependent on relief and has no other
food access, the full ration should be provided for all members of the population until they are able to re-
establish independent livelihoods. This situation is commonly found in sudden-onset disasters, such as
the early stages of displacement when beneficiaries may have lost all their assets and their usual sources
of food and income.

23.2. Targeting by family size

If a household is eligible for relief assistance, all members of the household should be registered and the
allocation of relief should be according to the full household size. The common practice of ‘capping’ the
number of beneficiaries per household reduces the impact of relief by diluting the benefit level. It is
unfair to larger households and is not acceptable.

23.3. Duration of relief eligibility

Relief assistance is provided to targeted weredas and communities for a temporary period, sufficient to
cover survival and recovery from a disaster. The expected duration of the need for relief is indicated by
the needs assessment, and the number of months of relief planned is included in the Humanitarian
Requirements calculations. The planned duration of relief distributions should be clearly communicated
to the beneficiary community at the beginning of the targeting process, and any changes in the planned
end date should also be immediately communicated to beneficiary communities via the wereda relief
committees.
Household and individual eligibility for relief is normally for the duration of the relief operation and will
end when the relief operation in their area ends: therefore no ‘graduation’ criteria are needed at these
levels, except in nutritionally-targeted supplementary or therapeutic feeding programmes which have
defined anthropometric thresholds for entry and exit (see Section 24 below, and ENCU technical
guidelines on nutrition interventions). Communities may, with guidance from the wereda and kebele
committees, allocate varying numbers of months’ support to different groups of households according to
need, at the time of registration. Once the registration list is agreed, beneficiaries should know how
much aid they can expect to receive, when and for how long.
Appeals procedures (see section 20.4.a.) allow for minor adjustments to beneficiary registration lists to
correct errors of inclusion or exclusion. Major re-targeting, involving a repetition of the whole
community targeting process set out above, should only be necessary if the quantity of aid allocated to
the community is significantly changed. Beneficiary lists should not be rotated in order to spread aid
among a larger number of people: this is a major cause of benefit dilution.

24. Supplementary foods (SF)

24.1. Nutritional targeting of curative SF (for acutely malnourished individuals)

The Targeted Supplementary Food (TSF) programme is implemented by DRMFSS and WFP under the
Extended Outreach Strategy of the Ministry of Health and UNICEF. It is a curative programme targeted at
individual level to women and children who are already acutely malnourished. Individuals are screened
by trained staff from the Ministry of Health, Health Extension Workers and community volunteers, using
anthropometric measurement of Mid-Upper-Arm Circumference (MUAC). Fortified blended food and
vegetable oil for TSF are provided by WFP through Regional DPPBs.

67
TARGETING RELIEF FOOD ASSISTANCE PART FIVE: BENEFICIARY TARGETING

Ad-hoc or emergency TSF may also be implemented in areas of acute need. Technical guidelines on
targeting procedures for nutritional interventions are provided and periodically updated by the
Emergency Nutrition Co-ordination Unit (ENCU) of the DRMFSS.

 See ENCU 2004, Emergency Nutrition Intervention Guideline (Updated version due 2011)

24.2. Categorical targeting of preventive SF or ‘relief CSB’ (for vulnerable groups)

Supplementary foods such as Corn-Soy Blend (CSB) or Famix are also sometimes provided alongside
general rations, as part of the mainstream relief system. In this case, the high-nutrient additional foods
are intended for physiologically vulnerable groups within the beneficiary community (namely children
under five and women who are pregnant or nursing), in order to prevent them becoming acutely
malnourished.
In the standard terminology of emergency nutrition, this kind of SF is referred to as ‘blanket
supplementary feeding’. This means it is intended for all members of the defined vulnerable groups, in
contrast to ‘targeted supplementary feeding’ which requires individual screening of nutritional status. It
does not mean that all relief beneficiaries are entitled to a share of the supplementary foods. Sharing
supplementary foods among the wider population is a waste of a scarce and valuable resource, and most
importantly deprives the targeted vulnerable groups of the needed additional nutrients.
Procedures for registering vulnerable group members for SF distributions may be adapted to the local
situation and capacities. If TSF-trained staff and volunteers are available in the locality it is recommended
that they be involved, as they will have a good understanding of the rules and reasons for SF targeting
and can help to educate community members on these issues as well as to implement the distribution. If
possible the wider community or their representatives should be asked to help in registering vulnerable
group members to ensure that no eligible individuals are excluded. However, in this case the community
is not empowered to adjust the targeting criteria.

DRM Principle 12: Give due attention to the most vulnerable groups
(see Table 3 in Part Two)

24.3. Co-ordinating GFD with Supplementary and Therapeutic Feeding

In areas where nutritionally-targeted interventions such as TSF or CMAM (Community Management of
Acute Malnutrition) are in operation, the households of individuals who are registered for such
nutritional support should automatically be included in the registration for GFD. That is, current acute
malnutrition should be an inclusion criterion for general rations. All international experience shows that
supplementary feeding is only effective when the household has adequate access to basic foods.
In PSNP weredas, similarly, no household should be excluded from nutritional programmes (e.g. TSF,
CMAM, or blanket supplementary feeding) on the grounds that they are already receiving support from
the PSNP. Overlaps between these programmes at the household level are desirable and represent
effective targeting of complementary assistance types.

68
TARGETING RELIEF FOOD ASSISTANCE PART FIVE: BENEFICIARY TARGETING

25. Co-ordinating relief with the Safety Net (for PSNP weredas only)
The objectives of co-ordinating the targeting of relief and PSNP transfers are, in order of priority:
1. To ensure continuity of assistance, so that no-one who needs emergency food assistance should ‘fall
through the gaps’ between PSNP instruments and relief.
2. To avoid confusion and duplication of efforts, by harmonising rules and procedures and making use of
existing capacities.
3. To exchange lessons learned and to build on the experience of both the PSNP and the emergency
relief system.

DRM Principle 6: Protection of human life

(see Table 3 in Part Two)

DRM Principle 8: Ensuring basic needs

(see Table 3 in Part Two)

25.1. Phased response to shocks: PSNP, Risk Financing and relief

Figure 1 in Part Two visualises the phased response to food insecurity in areas supported by the
Productive Safety Net Programme (PSNP): in normal times the safety net provides a predictable level of
support to chronically food insecure households. If a shock or disaster strikes, additional needs will be
met by temporarily scaling up the safety net, using first the PSNP regional Contingency Budget, and then
(if the scale or duration of needs exceeds the contingency resources) the federally-managed Risk
Financing Mechanism (RFM). The RFM, introduced in 2010, is expected to significantly reduce the size
and frequency of relief food operations in PSNP weredas. It is designed to provide a quicker and earlier
response than the relief system has sometimes been able to deliver in the past, and may provide useful
lessons for future relief operations.
If the scale or duration of emergency needs exceeds the resource ceiling of the combined Contingency
Budget and RFM, then the relief system will be required to step in. Since the PSNP and relief systems use
the same early warning and needs assessment process, there should be ample time to predict and plan
for relief needs in the PSNP weredas. This is particularly important when preparing a Humanitarian
Requirements Document (HRD). Ideally, the HRD and RFRD (Risk Financing Requirement Document – see
Section 17.1.c.) should be prepared at the same time, using the same information, so that together they
represent the total number of people in need. Only if the PSNP is unable to cover all the transitory-acute
needs will a wereda be provided with relief food assistance. However, other types of relief or emergency
services may be mobilised in PSNP weredas at any time according to need.

 See MoARD 2009, Final Guidelines for the PSNP Risk Financing Mechanism in Ethiopia

25.2. Household targeting of PSNP and relief assistance

When relief food assistance is to be distributed in a PSNP wereda, the targeting principles and broad
criteria set out in these guidelines will apply. That is, the needs and coping capacity of potential
beneficiary households should be assessed based on all their sources of food and income, including any
PSNP transfers they are receiving. If they cannot meet their short-term food needs without employing

69
TARGETING RELIEF FOOD ASSISTANCE PART FIVE: BENEFICIARY TARGETING

damaging coping strategies, they are eligible for relief whether or not they are PSNP clients. The
seasonality of PSNP transfers should also be taken into account: PSNP client households who are affected
by a shock or disaster may need additional assistance during the months when they do not receive
safety-net payments.

Box 6: PSNP rules for households affected by shocks

“When there is a shock such as a drought or flood, the PSNP is able to expand temporarily (or „scale-up‟) to
protect households which are affected by the shock. PSNP scales up to the level, and for the duration of time,
required to ensure that livelihoods are adequately protected. This allows the livelihoods of households that
are not PSNP clients to be temporarily protected. It also allows PSNP clients to gain additional
protection if needed. Such transfers follow the same rules for conditionality as transfers to chronically food
insecure households and are therefore categorised as either Conditional or Unconditional. The PSNP can
scale-up to the level of contingency budget and Risk Financing resources, any response beyond this comes
through the emergency response system.”
source: PSNP PIM (MoARD 2010:8, emphasis added)

Targeting and registration of households for relief food assistance in PSNP weredas should be carried out
by the existing PSNP Food Security Task Forces (FSTFs) at community, kebele and wereda level, with
additional support and guidance from relief staff of government and humanitarian partners if available.
When relief distributions are to be made in a PSNP wereda, a ‘single registry’ system should be
introduced, i.e. there should be a single registration list or database showing the amounts received by
each beneficiary household from PSNP and/or from relief.
If the PSNP is able to expand the public works component of the safety net sufficiently to absorb the
numbers of people requiring relief, then the conditionality rules of PSNP may be applied to the relief
transfers. However, if the scale of the transitory needs exceeds the capacity of the wereda to employ
beneficiaries in meaningful and appropriate work, or if there is a high prevalence of acute malnutrition or
illness among the temporary beneficiaries, it is recommended that relief assistance be provided
unconditionally for the period of the emergency. The PSNP principle of Primacy of Transfers states that
beneficiaries (clients) should receive the transfers they are entitled to in a timely manner, and that any
delays in administration or implementation of public works should not affect the timeliness of transfers.
This principle will be particularly important when dealing with acute needs.
Households who qualify for relief (i.e. they have been affected by a disaster and are unable to meet their
food needs in the short term without recourse to damaging coping strategies), but who are considered to
be chronically food-insecure (i.e. they are also unable to meet their food needs in the longer term, even
without a disaster event), should be treated as follows:
1st choice: If the PSNP is able to meet the household’s immediate needs through the CB or RFM, the
eligible household should be assisted through these mechanisms and then considered for registration in
the regular PSNP at the next re-targeting;
2nd choice: If immediate support from the PSNP is not available or is not sufficient to cover short-term
emergency needs, the household should be provided with relief assistance for the duration of the
disaster response and then considered for registration in the regular PSNP at the next re-targeting.
3rd choice: If no support from the PSNP is available to an eligible household in a disaster-affected area
(either because of resource constraints, or because the household does not reside in a PSNP kebele, or
for any other reason), then the household should be provided with relief assistance for the duration of
the disaster response. The emergency relief system cannot provide longer-term assistance to address
chronic problems.
No household should be excluded from relief assistance on the grounds that they are chronically food
insecure, unless they are already registered in the PSNP and receiving adequate support to cope with the

70
TARGETING RELIEF FOOD ASSISTANCE PART FIVE: BENEFICIARY TARGETING

additional emergency shock. In that case, the recommended household targeting process should find
that they are not eligible for relief, because the PSNP transfers are included in the assessment of
household food and income sources. The fundamental responsibility of the relief system is to save lives
and livelihoods in situations of urgent need.
The procedures for coordinating between relief and PSNP targeting should be reviewed and developed
jointly by the EWRD and FSCD as experience is gained with the implementation of both the Risk Financing
Mechanism and these revised National Guidelines on Targeting Relief Food Assistance.

26. Unconditionality of relief assistance

Under present government policy (in contrast to the previous NPDPM) there is no requirement for relief
food assistance to be linked to participation in employment schemes or community works. International
humanitarian norms require emergency relief to be provided unconditionally to people whose lives or
livelihoods are at risk. Previous experience in Ethiopia has also highlighted the practical difficulties of
organising successful asset-creating work schemes in an emergency context, when time and resources
are usually too scarce for proper planning and implementation and the duration of assistance is
unpredictable. Therefore, relief food assistance should by default be provided unconditionally according
to need.
In exceptional circumstances, Regional governments may authorise weredas to coordinate relief
distributions with community works or other previously planned and properly organised activities. In
these cases, the participation of relief beneficiaries in such activities should be subject to the following
strict conditions:
1. The timing of relief food distributions should never be dependent on the timing of works. Timing is a
key element of effective targeting, and relief should always be distributed when it is needed.
Weredas must not delay distributions for any reason other than an agreed evidence-based
assessment of when the relief is needed.
2. Only able-bodied, healthy adults should be required to participate in physical activities (see Box 7).

Box 7: People who should never be required to work in exchange for relief

 Children (under the age of 16)29

 Elderly people (over the age of 55, or those considered by the community to be unable to do a full
adult workload because of their age)
 People with physical or mental disabilities that prevent them working
 People temporarily unable to work, including:
o Women from the 6th month of pregnancy
o Nursing mothers in the first 10 months after giving birth
o People who are sick, acutely malnourished, or injured

These categories are in line with PSNP rules for participation in public works (2010 PIM, p.161).

3. Registration for emergency relief should not be connected to labour capacity, and no ‘quotas’ should
be set for the proportion of relief beneficiaries who are required to work. In a disaster-affected

29 The International Convention on the Rights of the Child defines children as those aged under 18. However, it may
be unrealistic to enforce this age limit in Ethiopia in the short term. The minimum age of 16 for employment on
community works is consistent with PSNP rules.

71
TARGETING RELIEF FOOD ASSISTANCE PART FIVE: BENEFICIARY TARGETING

population, health and nutritional status may be compromised and for various reasons there may be
a higher proportion of vulnerable group members than usual. Everyone who needs relief and is
eligible according to the locally-agreed criteria should be included in the registration, regardless of
their ability to work.
4. Household allocations of relief rations should be based only on need and family size, not on the
amount of labour contributed.
5. The timing and quantity of work required should be agreed in consultation with participants and the
wider community, to ensure that it does not disrupt or conflict with essential livelihood activities.
This consultation process should include separate discussions with women participants to ensure
that their total workload is not excessive, and that the works do not conflict with their domestic and
childcare responsibilities.

27. Adapting targeting processes and criteria to local contexts

Details of community targeting processes and household targeting criteria need to be locally adapted in
respect of variations in the following:
 livelihoods (drawing on the wealth of information in the HEA baselines);
 cultural and social factors (including household size and organisation; gender roles; extended family
or clan networks; religious values; local institutions for social support and redistribution; and
community perceptions of fairness in targeting);
 power structures and governance (in relation to the types of community targeting bodies and
processes likely to be effective); and
 the capacity of local government and communities to manage and record targeting and distributions.
Factors which tend to increase this capacity include prior experience of relief management, the field
presence of humanitarian partner organisations, and inclusion in the PSNP.
Regional authorities should take account of all these factors in developing detailed implementation
instructions and handbooks for the weredas.
Adapting the targeting process and criteria for pastoralist contexts may be particularly challenging, for
reasons which include population mobility, the tendency to share and redistribute resources within clan
structures (rather than the smaller more autonomous household structures found among settled
farmers), and livestock ownership systems which can make it difficult to identify appropriate asset-based
criteria. In addition to these inherent features, there has historically been a lack of experience with
targeted relief in many of Ethiopia’s pastoralist areas, and a lack of attention to their distinctive
traditional institutions and social structures. For example, the previous Food Aid Targeting Guidelines
were based on experience in the sedentary farming areas and did not give any special attention to the
different conditions in pastoralist societies.
Pastoralist livelihoods and societies are, however, very varied. There is no ‘one-size-fits-all’ targeting
system which will be appropriate for all pastoralist areas. Therefore, Regional authorities should base
detailed targeting instructions and handbooks on local knowledge and consultation with communities.
Regions have considerable flexibility to determine appropriate targeting processes and structures for
their populations, within the core principles and overall framework set out in these guidelines.

28. Monitoring for improved targeting

Monitoring of the targeting process and outcomes is a joint responsibility of government and partner
organisations at all levels. Section 15 in the Toolbox sets out some provisional checklists, indicators and
methods for the monitoring of relief targeting. These should be adapted or added to as needed,
according to local situations and the available resources from government and partners, to ensure
adequate monitoring of relief targeting in all Regions to verify that relief is reaching the right people and
that the targeting system is transparent and accountable.

72
TARGETING RELIEF FOOD ASSISTANCE PART FIVE: BENEFICIARY TARGETING

SPHERE Key Indicator 4:

“Distribution systems are monitored to ensure that targeting criteria are respected and that
timely corrective action is taken when necessary.”
(see Box 3 in Part Two)

If resources allow, a combination of corrective and analytical monitoring should be aimed for. This could
include regular representative surveys or consultations (for analysis of trends and impacts, as well as a
longer-term view of the targeting system), informal spot-checks (e.g. talking to people at distribution
sites or community meetings) and audits of registration lists, ration cards and appeals records (i.e. asking
to see the records at kebele or community level, checking any anomalies and discussing them with the
relevant people). Participatory monitoring methods are encouraged, to solicit the views and feedback of
community members including beneficiaries and decision-makers. Designated staff of government and
partner organisations should be authorised, by the Regional authorities through the wereda relief
committees, to conduct such monitoring activities whenever possible and to report their findings to the
authorities for action.
The purpose of monitoring is not data collection: unless the information is communicated to the right
people and (when necessary) acted on, the resources and effort spent in gathering it will be largely
wasted. Information flows are therefore crucial. In laying down the detailed lines of responsibility and
accountability for targeting, each Region should agree with partners on the exchange of monitoring
information (along the general lines of information flow in Figure 14 above), and designate individuals or
offices responsible for receiving monitoring information (whether formal reports or informal briefings)
and acting on it promptly when necessary. Regional Implementation Handbooks for these guidelines (see
Section 3.4) should include specific instructions to the weredas on monitoring and record-keeping.
The timely detection and correction of any abuses is essential for accountability, and is a key function of
monitoring. In order to recommend the appropriate action, monitoring should always seek to determine
the causes of any problems. Distinctions should be made between gross errors and marginal errors (see
Table 10 in Section 15.3), and between deliberate and unintended mis-targeting. Unintended mis-
targeting could be due to honest mistakes, disagreements, or inadequate understanding of the
guidelines. Deliberate abuse of targeting power would include nepotism, demanding bribes, and
favouring or excluding people on any basis other than need (e.g. personal connections, ethnicity, political
affiliation or religion). At the same time, monitoring should not be entirely focused on identifying
problems. This risks giving an unbalanced assessment of how well the targeting system overall is
operating. Therefore, best practices and successes should also be reported and learned from.
Lastly, monitoring visits are an opportunity to provide support and advice, and if appropriate to
recommend more substantial capacity development, for the decision-makers in local government and the
community. Local targeting committees, especially the unpaid representatives at community level, have
an extremely difficult and often thankless task. They need and deserve such support from outside the
community, in order to carry it out.

73
TARGETING RELIEF FOOD ASSISTANCE PART SIX: CROSS-CUTTING ISSUES

PART SIX: CROSS-CUTTING ISSUES

A cross-cutting issue is one that cannot be treated adequately as a separate topic, because it affects (or
‘cuts across’) a number of other issues or categories of action. Cross-cutting issues should therefore be
actively considered at all stages, and in all component parts, of a relief operation. This part of the
guidelines discusses the three key cross-cutting issues of gender equity, child protection and chronic
illness or disability, and brings together the action points on these issues that are contained in earlier
sections.

29. Gender equity

Gender equity means fairness to both women and men, taking account of differences in their
vulnerabilities and capacities according to the fundamental principle of needs-based targeting. Gender
differences in the roles, characteristics and opportunities of women and men are socially and culturally
determined, and vary from place to place and from time to time. Targeting and monitoring processes for
relief food assistance should therefore incorporate a situation-specific analysis of gender differences.
Women and girls in Ethiopia are exposed to a complex of social and economic vulnerabilities which are
summarised in Box 8. On the other hand, some differences in vulnerability are determined by biological
sex: at key points in their lives (during pregnancy and breastfeeding) women are among the most
physiologically vulnerable groups because of increased nutritional needs.

Box 8: Gendered risks and vulnerabilities in Ethiopia

“Poor households typically face a range of risks, both economic, which in Ethiopia are often
environmentally driven, and social. ... Both economic risks and social risks are influenced by gender
dynamics and may have important differential impacts on men and women. For example, women
typically have lower levels of education; less access, ownership and control of productive assets;
less access to credit; and different social networks than men, leading to lower economic productivity
and income generation and weaker bargaining positions in the household. This leaves them
particularly vulnerable to economic shocks. Furthermore, social sources of vulnerability, which are
often as or more important barriers to sustainable livelihoods and general well-being than economic
shocks and stresses, also typically work to particularly disadvantage women. Not only do women
lack voice in national and community fora, but their power in the household is often as limited as
their time. Female-headed households, typically suffering from a deficit of labour, are among the
poorest groups in Ethiopia. Elderly women also appear to be especially vulnerable, particularly in
the context of HIV/AIDS, due to their high burden of care.”
Jones et al. 2010: viii-ix

For all these reasons, special attention is needed to ensure that women receive fair treatment, including
an effective voice in decision-making, during the targeting of relief food assistance. However, it should
not automatically be assumed that all women are vulnerable or food-insecure. Although a female-
headed household is statistically more likely to be poor and vulnerable (in most societies, not only in
Ethiopia), this does not mean that all female-headed households are poor, or that male-headed
households are always more food-secure and resilient. Many women (though not yet many enough) are
also influential and outspoken leaders in their community. Gender is an important factor influencing
livelihoods, vulnerability and coping capacity, but other factors must also be taken into account to make a

74
TARGETING RELIEF FOOD ASSISTANCE PART SIX: CROSS-CUTTING ISSUES

fair assessment of each household’s need for relief assistance. No single indicator should be used to
target households in need of relief food assistance (see Section 21.1).
These guidelines encourage due attention to the needs and capacities of women in the following key
roles:
 Women as vulnerable group members (Sections 21 and 24). In setting criteria and registering
households, communities should pay due attention to the social, economic and physiological
vulnerabilities of women and girls. Observing the targeting rules on supplementary feeding for
vulnerable groups will help to ensure that women’s nutritional needs are met during the critical
periods of pregnancy and breastfeeding.
 Women as decision-makers and stakeholders in community-administrative targeting (Section 20).
Equal representation of women and men in relief committees and decision-making meetings at all
levels should be aimed for: if this is not achievable in the short-term, measures should be taken to
ensure at least some representation of women’s views and interests in the decision-making process.
When community representatives are elected, facilitators should ensure there are women
candidates as well as men. If it is difficult for women to participate fully in mixed meetings, separate
women’s meetings should be encouraged.
 Women as sources of knowledge (Sections 21 and 28). Women are experts on food security and
food management, and often know more than anyone else about who is hungry in their community
and how people are coping with acute food shortages. Assessment and monitoring missions should
specifically seek to interview women as key informants or in focus groups. Community targeting
processes should ensure that women’s opinions on criteria and household selection are heard and
valued.
 Women as household food managers (Section 22). Registering women as the ‘food entitlement
holders’ for their household and issuing ration cards in their name supports their dignity and
authority as household food managers. It also helps to ensure that food assistance reaches children
and other dependents, and that potentially vulnerable groups are not excluded if male household
heads are absent.
The following institutional and information-generating elements of the relief targeting system should also
be improved in the interests of gender equity:
 Women-friendly appeals (Section 20.4.a.). Because women often lack a voice in local power
structures and may be intimidated from coming forward with appeals or complaints, or their appeals
may be less likely to succeed,30 it is recommended that there should be at least one woman member
in the kebele appeals body who has special responsibility to receive and follow up appeals from
women. Authorised government or partner staff engaged in spot-checks, audit or regular monitoring
(Section 28) should also ensure that women have access to them to report any problems or
complaints about the targeting process.
 Empowering women’s representatives in local government (Section 20.1.). The heads of the wereda
Women’s Affairs Department and the kebele Women’s Association should be given special
responsibility, together with training and resources if possible, to represent the interests of women
in relief targeting.
 Monitoring and record-keeping to enhance gender equity (Sections 20.4.c. and 28). Information is
power: without reliable gender-disaggregated records of registration lists, appeals, and other aspects
of the targeting process, it is extremely difficult to know how well the relief system is addressing the
different needs of women and men, or what actions might be needed to improve it.

30 See Sharp et al. (2006) for evidence of these problems in the similar targeting system of the PSNP.

75
TARGETING RELIEF FOOD ASSISTANCE PART SIX: CROSS-CUTTING ISSUES

 See FDPPC 2004, Gender Mainstreaming Guidelines For Disaster Management Activities

It is recommended that the Gender Mainstreaming Guidelines for Disaster Management Activities should
be reviewed and updated, and should be included in the portfolio of revised guidelines being compiled by
DRMFSS.

30. Child protection

The Sphere Minimum Standards in Disaster Response state that “special measures must be taken to
ensure the protection from harm of all children and their equitable access to basic services”, noting that
“children often form the larger part of an affected population” in emergency situations (Sphere Project
2004:10). The following components of these guidelines are particularly relevant to child protection:
 Supplementary feeding (Section 24). Children under five years old are the demographic group most
vulnerable to malnutrition in emergencies. Severe malnutrition in this age group not only causes
high rates of illness and death, but can impair long-term physical and mental development.
Observing the targeting rules on supplementary feeding for vulnerable groups will help to ensure
that their rights to adequate food and future health are protected.
 Child labour (Section 26). Children under 16 should not be employed on community projects or
public works, whether through the PSNP or other projects which may occasionally be linked to relief.
If a household which is registered for relief assistance does not have enough able-bodied adult labour
to meet a work requirement without sending children, then relief should be provided to that
household unconditionally or with a reduced work requirement.
 Coping strategies (Sections 20.2 and 21). Targeting assistance to households who would otherwise
need to depend on unacceptable or damaging coping strategies will help to protect children from
harmful practices such as child labour, prostitution or begging, and withdrawal from school.
 Child-headed households (Section 31) should automatically be considered among the most
vulnerable groups for relief targeting.
 Full-family targeting (Section 23.2) will ensure adequate relief food supplies for children in larger
households.
 Respecting women’s childcare responsibilities (Sections 20.1.c. and 26). Women should be
consulted about the timing and workload implications of their participation in the community
targeting process and in any public works linked to relief. This will not only facilitate stronger
representation of women in decision-making and protect women from excessive workloads, but will
also help to protect the quality of care they can give their children.
 Registration of women as food entitlement holders (Section 22) will ensure that households
containing children are not unintentionally excluded from relief targeting in situations where the
male household head may be absent. This includes the common scenario in pastoralist livelihood
zones where the men have temporarily migrated with the livestock, leaving the women, children and
elderly.

31. HIV/AIDS, chronic illness and disability

People living with HIV/AIDS (PLWHA) and other major illnesses such as tuberculosis and malaria are
vulnerable in food emergencies because they have higher nutritional needs, may be socially marginalised,
are less likely to be able to cope with shocks by their own efforts, and may already have eroded their
assets because of the illness. Adult deaths from these diseases also leave households more vulnerable to
food insecurity. Not only have the survivors lost a potential breadwinner, but they are also likely to have

76
TARGETING RELIEF FOOD ASSISTANCE PART SIX: CROSS-CUTTING ISSUES

depleted their assets and compromised their own health caring for the deceased during long periods of
illness. Both their long-term food security and their capacity to cope with a shock are likely to be
damaged. Research suggests that HIV/AIDS deaths in parts of rural Ethiopia may be a significant cause of
the high number of impoverished female-headed households (Uraguchi 2010: 493). Elderly-headed
households caring for orphaned children and child-headed households are other common types of
HIV/AIDS-survivor household.
Targeting and distribution processes should not expose people to HIV infection risks (for example, by
requiring them to spend a night in town or far from their homes waiting for a distribution). Adequate,
timely and well-targeted relief food assistance should also prevent disaster-affected people from needing
to resort to distress strategies that expose them to risks of HIV infection (such as prostitution). This is a
particular danger for women and for young people, and should be considered during the community
discussion of acceptable and unacceptable coping strategies.
Targeting processes and criteria should respect the dignity of disaster-affected people and should not
impose stigma or shame on any potential beneficiaries, including those who are chronically ill or their
dependents and survivors. Equitable consideration of people’s vulnerability in terms of their coping
capacity and resources, including their ability to work, should ensure that people who are chronically ill
(whether from HIV/AIDS or other debilitating illnesses such as malaria and tuberculosis) are included.
Medical diagnosis of HIV/AIDS or other illness is not necessary for a household to be eligible for relief:
community members know when individuals are unable to work, or a household lacks enough working
adult members to support itself, or is spending large amounts of time and money caring for sick
members. The proxy indicators in Box 9 can be used to ensure inclusion of households affected by
HIV/AIDS and other chronic illnesses. Actual diagnosis of HIV-positive status should only be used as a
targeting criterion when patients are enrolled in a medical treatment programme and food is provided in
support of the treatment.

Box 9: Proxy indicators for households affected by HIV/AIDS

 Labour-poor households (lacking enough able-bodied labour to earn adequate income for the household)
 Elderly-headed households (household head > 60 years old) with children and little or no adult labour
 Child-headed households (household head < 16 years old)
 Households hosting orphans and other vulnerable children
 Households with chronically ill working-age adults, or who have suffered the death of working-age adults.

Care should be taken to ensure that people with disabilities are represented in decision-making processes
for relief targeting, that their needs and vulnerabilities are equitably assessed, and that relief is
distributed in a way that is accessible to them. As the Sphere Handbook (2004:11) states, “In any
disaster, disabled people - who can be defined as those who have physical, social or emotional
impairments which make it difficult for them to use standard disaster response services – are particularly
vulnerable”. No-one should be excluded from relief assistance because of stigma or marginalisation
connected with either illness or disability.

77
TARGETING RELIEF FOOD ASSISTANCE GLOSSARY

GLOSSARY

Acute malnutrition Protein-energy malnutrition caused by a recent and severe lack of food intake or
disease that has led to substantial weight loss or nutritional oedema. There are
different degrees of acute malnutrition, which are often categorized as follows:
moderate acute malnutrition (MAM) corresponding to –3 to <–2 Z scores or 70 to 80%
median weight-for-height;
severe acute malnutrition (SAM) corresponding to <–3 Z scores or <70% median
weight-for-height and/or nutritional oedema; and
global acute malnutrition (GAM) encompassing both of the above and corresponding
to <–2 Z scores or <80% median weight-for-height and/or nutritional oedema.
Sometimes called Total Acute Malnutrition (WFP EFSA).
The GAM rate is a standardised indicator of the level of acute food insecurity (&/or
health crisis) in a community or area, and is therefore used in geographical targeting
and “hot-spot” prioritisation.

Affected population People whose lives or livelihoods are affected by a disaster. In many disasters,
particularly in slow-onset crises, not everyone who is affected will be in need of relief
food assistance. This will depend on their vulnerability and their access to coping
opportunities.

Beneficiaries People who benefit from a relief distribution, i.e. those who actually receive the aid.
(Relief operations may also have indirect beneficiaries, who do not receive assistance
themselves but whose lives are improved in other ways: however, for targeting
purposes, the focus is on direct beneficiaries).

Blanket distribution A distribution that covers everyone in a targeted area or group, without further
selection criteria.
This term can be confusing because it is used at different targeting levels: it does not
mean there is no targeting. For example, „blanket supplementary feeding‟ provides
rations to everyone in the targeted vulnerable groups, usually children under five and
women who are pregnant or lactating. By contrast, „targeted supplementary feeding‟
uses anthropometric screening to select malnourished individuals within these target
groups.

Capacity The combination of all the strengths, attributes and resources available within a
community, society or organization that can be used to achieve agreed goals.
(UNISDR)

Community A group of people with a commonality of association, having common interest, shared
experience, or function and living in the same locality and under the same public
administration. (DRM policy)

Complex emergency A humanitarian crisis in a country, region or area where there is a total or considerable
breakdown of authority resulting from internal or external conflict, and which requires
an international response that goes beyond the mandate or capacity of any single
agency and/or the on-going UN country programme.
(WFP 2005b)

78
TARGETING RELIEF FOOD ASSISTANCE GLOSSARY

Coping capacity The capacity of households to diversify and expand access to sources of food and
income, and thus to address deficits created by a specified hazard. (Boudreau 2009)

Coping strategies Activities that people resort to in order to obtain food, income and/or services when
their normal means of livelihood have been disrupted. (WFP 2005b)
Acceptable coping strategies:
“Coping strategies that are sustainable and preserve future means of livelihood,
dignity and nutritional health.”
Unacceptable / damaging coping strategies:
“Strategies that undermine future means of livelihood, dignity or nutritional health,
increase long-term vulnerability, or are illegal or not socially acceptable.” (WFP
EFSA)

Coverage Successful provision of assistance to the intended target group.

Quantitative measure: the percentage of the intended or eligible target group who
actually receive assistance.

Destitution “a state of extreme poverty that results from the pursuit of „unsustainable livelihoods‟,
meaning that a series of livelihood shocks and/or negative trends or processes erodes
the asset base of already poor and vulnerable households until they are no longer
able to meet their minimum subsistence needs, they lack access to the key productive
assets needed to escape from poverty, and they become dependent on public and/or
private transfers.” (Devereux 2003:11)

Dilution A common targeting problem in which resources are spread too thinly among too
many people, so that each beneficiary receives less than the planned amount of aid.
Dilution is the opposite of concentrating scarce resources where they will have the
greatest impact, which is the purpose of targeting.

Disaster The occurrence of an abnormal event triggered by a hazard that impacts a given entity
(including a vulnerable community, geographical area, infrastructure or a group of
people in significant numbers gathered in one place) causing damage and/or
casualties, leaving the affected entity unable to function normally, eroding survival
mechanism and forcing it to seek outside assistance for survival.
(DRM Policy FDRE 2010a)

Disaster Risk Collective actions and efforts of concerned institutions, policies, programs, and other
Management (DRM): measures designed to prevent, mitigate, prepare for, and respond to a disaster and to
provide recovery and rehabilitation support.
DRM aims to lessen the impacts of disasters, as well as reducing their frequency
through DRR.

Disaster Risk Reduction The broad development and application of policies, strategies, and practices to
(DRR) minimize vulnerabilities throughout communities via prevention, mitigation, and
preparedness. (DRM policy FDRE 2010a)
DRR addresses the causes of disasters.

Displacement The process of people being forced to move from their homes to other places because
of a natural hazard, war/conflict, or other human-made action. (DRM policy FDRE
2010a)

Emergency An urgent situation in which there is clear evidence that an event or series of events

79
TARGETING RELIEF FOOD ASSISTANCE GLOSSARY

has occurred which causes human suffering or imminently threatens human lives or
livelihoods. (WFP 2005a)

Emergency response The provision of essential goods and services to a disaster affected population to save
lives and protect livelihoods in times of disaster. (DRM Policy FDRE 2010a)

Exclusion error People who are wrongly excluded from food assistance. That is, people who are
eligible according to the targeting definition or criteria but who do not receive any
assistance.
Quantitative measure: the percentage of the intended target group who are not
beneficiaries.

Famine Widespread starvation; a period when the death rate in a population rises because of
causes related to food shortage or undernutrition.

Fast-onset (or sudden- A calamity which strikes with little or no warning and has an immediate adverse impact
onset) disaster on human populations, activities and economic systems.
This includes both sudden natural disasters such as floods, earthquakes, landslides
and cyclones, and human-made crises such as conflict and forced population
displacement. (WFP 2005a)

Food access (at A household‟s ability to regularly acquire adequate amounts of food through a
household level) combination of their own home production and stocks, purchases, barter, gifts,
borrowing or food aid. (WFP 2005b)

Food access shortfall (at The difference between households‟ nutritional requirements – what they need in
household level) order to re-establish or maintain satisfactory nutritional health and to carry out
productive activities – and what they are able to provide for themselves without
adopting distress strategies. (WFP 2005b)

Food availability The amount of food that is physically present in a country or area through all forms of
domestic production, commercial imports and food aid. (WFP 2005b)

Food assistance „Food assistance‟ is a broader term than „food aid‟. Definitions vary (see Harvey et al.
2010), but WFP‟s is:
“the set of instruments used to address the food needs of vulnerable people....
[including] in-kind food aid, vouchers and cash transfers.” When food assistance is
provided in kind, it may be either imported or locally purchased.

Food insecurity A situation “when people lack secure access to sufficient amounts of safe and
nutritious food for normal growth and development, and an active and healthy life.
Food insecurity may be caused by the unavailability of food, insufficient purchasing
power, inappropriate distribution, or inadequate use of food at the household level.”
(WFP Evaluation Glossary)

Food security Food security exists when all people, at all times, have physical and economic access
to sufficient safe and nutritious food for a healthy and active life.

Gender The roles and characteristics of women and men which are socially and culturally
determined.

Geographical targeting The selection or prioritisation of an area to receive assistance. Usually this is followed
by further levels of targeting to select groups, households or individuals within the
area.

80
TARGETING RELIEF FOOD ASSISTANCE GLOSSARY

GFD General Food Distribution (≠ SF or Supplementary Foods) – the distribution of a basic

staple ration providing the minimum nutritional requirements, mainly energy
(kilocalories) and protein.

Hazard A potentially damaging physical event, phenomenon, or human activity that may
cause the loss of life or injury, property damage, social and economic disruption, or
environmental degradation. (DRM Policy / UNISDR)

Hunger The experience (physical and psychological) of not having enough to eat.

Inclusion error People who are wrongly included in food assistance. That is, people who receive
assistance but do not meet the targeting definition or criteria.
Quantitative measure: the percentage of actual beneficiaries who do not meet the
targeting criteria.

Livelihood Zones “Geographical areas within which people share broadly the same patterns of access to
food and income and the same access to markets, thus making them vulnerable to the
same hazards”. (Boudreau 2009: v)

Malnutrition “Impairment of physical and/or mental health resulting from a failure to fulfil nutrient
requirements. Malnutrition may result from consuming too little food, a shortage of
key nutrients, or impaired absorption or metabolism due to disease.”
“Severe malnutrition results in very high infant and child mortality and, for those
children who survive, there are many life-long medical complications, including mental
retardation. Research has also demonstrated that even mild-to-moderate malnutrition
significantly raises the risk of mortality in children.” (USAID)

Mid-upper arm The circumference of the upper arm measured at the mid-point between the shoulder
circumference (MUAC) and the elbow, which is an approximate indicator of wasting in children 6 to 59 months
of age and pregnant women.
MUAC is typically used for rapid assessment and screening for acute malnutrition in
emergency situations.

Micronutrients Micronutrients include all vitamins and minerals essential for a wide range of body
functions and processes. (WFP EFSA)

Non-food needs Basic “household needs apart from food: in particular, shelter, fuel, cooking utensils,
water, health care, basic education and personal security.” (WFP EFSA)

Nutritional oedema Swelling (fluid retention) caused by malnutrition.

“Oedema results from the excessive accumulation of extracellular fluid as a result of
severe nutritional deficiencies, and is a serious cause for concern. All children with
nutritional oedema are automatically classified as severely malnourished.” (Young
and Jaspars 2006:4)

Nutritional requirements The amount of energy, protein, fat and micronutrients needed for an individual to
sustain an active and healthy life. (WFP EFSA)

Pastoralist People whose livelihood depends primarily on livestock.

Proxy indicator A substitute or indirect indicator which is statistically associated with a characteristic
that is more complex or difficult to measure, and is therefore used to represent it.

81
TARGETING RELIEF FOOD ASSISTANCE GLOSSARY

Recovery “actions taken after a disaster to restore the living conditions of disaster-stricken
people or communities while encouraging and facilitating necessary adjustments to
build resilience to prevent similar crises in future.” (DRM Policy FDRE 2010a)

Rehabilitation “the measures applied after a disaster which are necessary to restore normal activities
and build resilience to future shocks in affected areas, communities, and economic
sectors”. (DRM Policy FDRE 2010a)

Relief “Relief is defined as action that is intended to save lives and reduce suffering”
(ALNAP)
“...activity undertaken in the immediate aftermath of a disaster to save lives, protect
livelihoods, and address immediate humanitarian needs, including the provisional
restoration of essential services”. (DRM Policy FDRE 2010a)

Resilience “The capacity of a system, community or society potentially exposed to hazards to

adapt, by resisting or changing in order to reach and maintain an acceptable level of
functioning and structure. This is determined by the degree to which the social
system is capable of organising itself to increase this capacity for learning from past
disasters for better future protection and to improve risk reduction measures.”
UN/ISDR. Geneva 2004. Hyogo Framework for Action

Risk “the probability of a specific hazard occurring in a given location and its likely
consequences for people and property” (DRM Policy FDRE 2010a)

Slow-onset disasters Critical situations that develop slowly over time including natural disasters such as
drought, crop failures, pests, diseases and economic crises that result in an erosion of
families‟ capacities to meet their food needs. (WFP EFSA)

Starvation “life-threatening undernutrition, when a poor nutritional state unilaterally raises

vulnerability to mortality through any means” (De Waal 1991).

Supplementary Feeding Supplementary feeding (SF) is the provision of additional nutritious rations for those
with higher nutritional needs (such as pregnant women, lactating women with infants
under 6 months), or those who are already moderately malnourished. To be effective,
SF should be provided in addition to an adequate general ration, or in situations where
beneficiaries have adequate access to basic food.

Stunting (shortness) An indicator of chronic malnutrition. The prevalence of stunting reflects the long-term
nutritional situation of a population.
It is calculated by comparing the height-for-age of a child with a reference population
of well-nourished and healthy children. (WFP EFSA)

Therapeutic feeding Feeding and medical treatment to rehabilitate severely malnourished children.
(WFP EFSA)

Vulnerability “The conditions determined by physical, social, economic, and environmental factors
or processes, which increase the susceptibility of a community to the impact of
hazards”. UN/ISDR. Geneva 2004. Hyogo Framework for Action
The potential to suffer harm or loss. Determinants of vulnerability include physical,
social, economic, political, cultural, and institutional factors. (DRM Policy)
The characteristics and circumstances of a community, system or asset that make it
susceptible to the damaging effects of a hazard. (UNISDR)

82
TARGETING RELIEF FOOD ASSISTANCE GLOSSARY

Wasting (thinness) An indicator of acute malnutrition that reflects a recent and severe process that has
led to substantial weight loss. This is usually the result of starvation or disease and is
strongly related to mortality.
It is calculated by comparing the weight-for-height of a child with a reference
population of well-nourished and healthy children.

83
TARGETING RELIEF FOOD ASSISTANCE REFERENCES

REFERENCES AND FURTHER READING

Behnke, R., S. Devereux, A. Teshome, M. Wekesa and R. White (2007). Piloting the Productive Safety Net
Programme in Pastoral Areas of Ethiopia: Revised Programme Proposal. Addis Ababa, Food
Security Coordination Bureau and Pastoralist Task Force
van den Boogaard, R. (2006). Experiences of Targeting Resource Transfers and Interventions to Pastoral
and Agro-pastoral Communities. Addis Ababa, SC-UK / SC-USA Ethiopia
Boudreau, T. (2009). Solving the risk equation: People-centred disaster risk assessment in Ethiopia.
Network Paper 66. London, Humanitarian Practice Network, ODI
Devereux, S. (2003). Conceptualising Destitution. IDS Working Paper 216. Brighton, Institute of
Development Studies
Devereux, S. (2004). Targeting Social Protection. ppt Presentation (unpublished). Brighton, Institute of
Development Studies
Devereux, S. (2006). Distinguishing between chronic and transitory food insecurity in emergency needs
assessments, WFP / SENAC
DPPA (2007). Implementing Strategy for Verification Assessment. Addis Ababa, Early Warning
Department, DPPA in collaboration with the Methodology sub-group of the Early Warning
Working Group
DPPC (2000) National Food Aid Targeting Guidelines, Addis Ababa, Disaster Prevention and Preparedness
Commission
DPPC (2002) Food Aid Targeting Handbook, Addis Ababa, Disaster Prevention and Preparedness
Commission
Drèze, J. and A. Sen (1989). Hunger and Public Action. Oxford UK, Clarendon Press.
DRMFSS (2009). Wereda Disaster Risk Profiling Programme: Approach Paper. Addis Ababa, Ministry of
Agriculture
DRMFSS (2010a). Guidelines for Identification and Analysis of Areas Affected by Hazard (Draft 2). Addis
Ababa, Early Warning Working Group, DRMFSS
DRMFSS (2010b). Humanitarian Requirements Document - 2010. . Joint Government and Partners'
document, dated January. Addis Ababa, Disaster Risk Management and Food Security Sector,
Ministry of Agriculture and Rural Development (MoARD)
DRMFSS (2010c). Revised Early Warning Guidelines (Draft). Addis Ababa, Early Warning and Response
Directorate, Disaster Risk Management and Food Security Sector. Government of Ethiopia
DRMFSS (2011). Revised Humanitarian Requirements Document for South and Southeastern Parts of the
Country. Joint Government and Humanitarian Partners' document, dated April 12. Addis Ababa,
Disaster Risk Management and Food Security Sector, Ministry of Agriculture
ENCU (2004). Emergency Nutrition Intervention Guideline. Addis Ababa, Emergency Nutrition
Coordination Unit, Early Warning Department, Disaster Prevention and Preparedness
Commission
FDPPC (2004). Gender Mainstreaming Guidelines for Disaster Management Activities. Addis Ababa,
Federal Disaster Prevention and Preparedness Commission
FDRE (2010a). National Policy and Strategy on Disaster Risk Management Final (10th) draft. Addis Ababa,
Federal Democratic Republic of Ethiopia

84
TARGETING RELIEF FOOD ASSISTANCE REFERENCES

FDRE (2010b) Growth and Transformation Plan 2010/11 - 2014/15. Addis Ababa, Ministry of Finance and
Economic Development
GFDRR (2009). Disaster Risk Management Programs for Priority Countries, Global Facility for Disaster
Reduction and Recovery, UN-ISDR / World Bank
Harvey, P., K. Proudlock, E. Clay, B. Riley and S. Jaspars (2010). Food aid and food assistance in emergency
and transitional contexts - A review of current thinking. HPG Commissioned Report. London,
Humanitarian Policy Group, ODI
Jaspars, S. and D. Maxwell (2008). Targeting in Complex Emergencies: Somalia Country Case Study.
Medford, Massachusetts, Feinstein International Center, Tufts University
Jaspars, S. and H. Young (1995). General Food Distribution in Emergencies: from Nutritional Needs to
Political Priorities. Relief and Rehabilitation Network (RRN) Good Practice Review 3. London, ODI
Jones, N., Yisak Tafere and Tassew Woldehanna (2010). Gendered risks, poverty and vulnerability in
Ethiopia: To what extent is the Productive Safety Net Programme (PSNP) making a difference? .
London, ODI
LIU (2009). Seasonal Assessment Manual. Addis Ababa, Livelihoods Integration Unit, DRMFSS
Lundberg, M. K. A. and P. K. Diskin (1994). Targeting Assistance to the Poor and Food Insecure: A Review
of the Literature. MSU International Development Working Paper 47. East Lansing, USA, Dept. of
Agricultural Economics, Michigan State University
MoARD (2008). National Guidelines for Livestock Relief Interventions in Pastoralist Areas of Ethiopia.
Addis Ababa, Ministry of Agriculture and Rural Development
MoARD (2009). Final Guidelines for the PSNP Risk Financing Mechanism in Ethiopia. (document
developed by theIDLgroup). Addis Ababa, Ministry of Agriculture and Rural Development
MoARD (2010). Productive Safety Net Programme: Programme Implementation Manual (2010). Addis
Ababa, Ministry of Agriculture and Rural Development, Government of the Federal Democratic
Republic of Ethiopia
Sandford, S. and Yohannes Habtu (2000). Emergency Response Interventions in Pastoral Areas of
Ethiopia. Addis Ababa, report commissioned by DFID
SC-US (2005). Targeting Beneficiaries: Somali Region, Ethiopia, Save the Children USA/EFO
Sharp, K., T. Brown and Amdissa Teshome (2006). Targeting Ethiopia's Productive Safety Net Programme.
Report for DFID Ethiopia, ODI (London), theIDLgroup (Bristol), A-Z Consult (Addis Ababa).
http://www.odi.org.uk/plag/resources/reports/psnp_targeting.pdf
Sphere Project (2004). Humanitarian Charter and Minimum Standards in Disaster Response. Geneva,
Sphere Project / Oxfam Publishing.
Taylor, A. and J. Seaman (2004). Targeting Food Aid in Emergencies. ENN Special Supplement 1. Oxford,
Emergency Nutrition Network. www.ennonline.net
TGE (1993). National Policy on Disaster Prevention and Management. Addis Ababa, Transitional
Government of Ethiopia
UNISDR (2007). Hyogo Framework for Action 2005 - 2015: Building the Resilience of Nations and
Communities to Disasters, United Nations International Strategy for Disaster Reduction.
www.unisdr.org/hfa
Uraguchi, Z. B. (2010). "Food price hikes, food security, and gender equality: assessing the roles and
vulnerability of women in households of Bangladesh and Ethiopia." Gender and Development
18(3): 491-501.
WFP (2005a). Definition of Emergencies. Policy Issues WFP/EB.1/2005/4-A/Rev.1. Rome, World Food
Programme

85
TARGETING RELIEF FOOD ASSISTANCE REFERENCES

WFP (2005b). Emergency Food Security Assessment Handbook. Rome, World Food Programme
WFP (2005c). Full Report of the Thematic Review of Targeting in WFP Relief Operations Ref.
OEDE/2006/1. Rome, WFP Office of Evaluation.
http://www.wfp.org/operations/evaluation/doclist.asp?section=5&sub_section=8&EBDoc=WFP/
EB.1/2006/7-B&Year=2006&ID=WFP086129
WFP (2006). Targeting in Emergencies. Policy Issues WFP/EB.1/2006/5 A. Rome, WFP.
http://www.wfp.org/eb/docs/2005/wfp043676~2.pdf
Young, H. and S. Jaspars (2006). The meaning and measurement of acute malnutrition in emergencies. A
primer for decision-makers. HPN Network Paper 26. London, Humanitarian Practice Network,
ODI

86