Choosing the response and targeting

1Response modalities

Several modalities exist, often combined:

• In-kind food distribution: useful when markets are not functioning or prices are soaring; basket calculated to cover about 2,100 kcal per person per day for a full ration.
• Cash transfers (cash, mobile money): more choice and dignity, support the local economy, if markets work and risks are managed.
• Vouchers (value or commodity): steer consumption towards certain foods or traders.
• Livelihoods support: seeds and tools, animal health, destocking, cash for work, small businesses, to rebuild self-reliance.

2Market analysis and choice of modality

Before choosing, analyse: are markets functioning? Can traders respond to increased demand without driving prices up? Do households prefer food or cash? Are the risks (security, diversion, protection) manageable? Are financial service providers available?

Tools such as price monitoring, rapid market assessment or the minimum expenditure basket (MEB) are used to set transfer values. The choice is reviewed regularly: a modality suited to month 1 may no longer be suitable in month 6.

3Targeting

Targeting determines who receives assistance. It often combines geographic targeting (areas in IPC 3 or above) followed by household targeting according to vulnerability criteria: consumption, income, assets, household size, presence of older, sick or disabled people, female- or child-headed households.

Methods: community-based targeting, household survey with a score, or a combination. The two errors to minimise: exclusion of vulnerable households and inclusion of non-vulnerable households. Criteria are explained publicly, lists verified, and a complaints mechanism makes it possible to correct errors.

4Protection, gender and accountability in the response

A food response can create risks: long walks to sites, theft, sexual exploitation in exchange for a place on the list, tensions between groups. Basic measures: nearby and safe sites, suitable timings, separate queues if necessary, staff trained in PSEA, assistance for people with reduced mobility.

Accountability requires information (who is entitled to what, when, how to complain), participation of communities, including women, in choosing criteria and modalities, and an accessible and confidential complaints mechanism.

5Response analysis: a six-step method

Response analysis is the link between assessment and programme design. It answers the question: of all the possible options, which one (or which combination) best meets the needs identified, in this context, with these risks and these resources? It is too often skipped, with the modality chosen out of organisational habit or donor preference. The Sphere standards (food assistance standard 6 and the livelihoods standards) and CaLP's Minimum Standard for Market Analysis (MISMA) require the choice to be justified and documented.

Step 1: start from objectives, not modalities

First, set out what the response must achieve: cover the food gap of 4,200 households over a five-month lean season; prevent the sale of the core breeding stock; enable sowing for the next season. Each objective can be achieved through several options. If you start with "we will do a food distribution", you rule out any comparison.

Step 2: list the options

For each objective, list all the technically feasible options: in-kind food, unconditional cash, value or commodity vouchers, cash for work, seed fairs, direct seed distribution, animal feed, veterinary care, commercial destocking, support to traders, integration into an existing social safety net.

Step 3: apply the feasibility criteria

Step 4: compare costs

Two measures are commonly used. The cost-transfer ratio compares the total cost (procurement, transport, storage, staff, payment fees) with the value received by the household. The cost-effectiveness indicator compares the total cost with an outcome (for example the cost per household moving from poor to acceptable consumption). WFP also uses the Omega value, which compares the nutritional value obtained per unit of cost across modalities. Comparisons must include hidden costs: warehouse losses, guarding, and the cost to the household of transport to the site or of reselling part of the ration at a loss.

Step 5: decide, document and plan a fallback

The decision is recorded in a two- to three-page note signed by the project manager and the head of programmes. It explains the options that were rejected and why, as well as the conditions that would trigger a change of modality (for example a price rise of more than 25% or the closure of the main market). A fallback plan is built in from the design stage: flexible budget, framework agreement with a food supplier, agreement with a second payment service provider.

Step 6: review

The response analysis is reviewed at least every three months, and immediately if a monitoring indicator crosses an agreed threshold. The modality that was appropriate in month 1 may no longer be so in month 6: markets recover, prices stabilise, preferences change.

The question a donor asks when reading a proposal: "Why this modality, here, now?" The answer must fit into a few sentences backed by data: market conditions, measured preferences, risks and comparative cost. A phrase such as "cash enhances dignity" is not enough on its own.

6Analysing markets for decision-making: tools and indicators

Market analysis for decision-making is not a full economic study: it seeks to answer a few precise questions within two to three weeks at most. Several tools exist, to be chosen according to the time and expertise available.

Trader survey questions

• Which goods do you sell and in what quantities per week? What is your current stock?
• If demand doubled, could you keep up? Within how many days? What would prevent you (credit, transport, security, licences)?
• Where do your goods come from? How many suppliers do you have? Are the roads open?
• Do you have access to credit from your suppliers? Do you accept mobile money?
• How have your purchase and selling prices changed since the shock?
• How many traders sell the same goods in this market? (risk of price collusion)

Interpreting supply elasticity

The decisive criterion is the capacity of supply to respond to the cash injection. Practical benchmark: if a majority of the traders interviewed say they could increase their supply by at least 50% in less than two weeks, with several suppliers and open roads, and if the planned injection represents a modest share of the volume usually traded, the risk of induced inflation is low. If the market is controlled by two or three wholesalers, if supply routes are cut or if the injection represents a large share of sales, mitigation measures are needed (support to traders, vouchers with negotiated prices, staggered payments) or another modality.

Calculating the weight of the injection

Example: Doumbaré market sells around 180 tonnes of cereals per month. The programme plans transfers to 3,400 households, whose cereal component corresponds to 45 kg per household per month, i.e. around 153 tonnes of potential additional demand. Even taking into account that households were already buying some of these cereals, the increase is very large compared with the usual volume: the capacity of wholesalers in the regional capital must be checked, payments staggered and prices monitored weekly during the first two months. This simple calculation avoids costly mistakes.

Market analysis is also needed for in-kind food: buying large quantities locally can push up prices for non-assisted households. WFP and FAO assess this risk before local and regional procurement.

7Targeting in practice: methods, calculating errors and verification

Targeting is not an administrative formality: it is the most political and most contested decision in a food security programme. WFP's guidance on targeting and prioritisation, the joint UNHCR-WFP guidance on targeting in refugee operations and Sphere standard 6.3 (targeting, distribution and delivery) set out the principles: vulnerability-based criteria, transparency, community participation, verification and a complaints mechanism.

Methods and conditions for their use

In practice, methods are combined: geographical targeting on the most severely affected areas, then community-based targeting using criteria derived from the assessment, verified by visits to a sample of households, with automatic inclusion criteria for a few highly vulnerable categories.

Measuring targeting errors

To measure errors, "targeted or not" status is compared with "vulnerable or not" status measured independently (for example a CARI category or the FCS) on a sample of households from both groups.

• Inclusion error rate = B ÷ (A + B) = 90 ÷ 400 = 22.5% of those registered are not vulnerable.
• Exclusion error rate = C ÷ (A + C) = 70 ÷ 380 = 18.4% of vulnerable households are not registered.

Caution: the sample of non-registered households must be drawn from the non-registered population, which requires a list or census of households. The rates also depend on the definition of "vulnerable": always state it.

There is no universal threshold for an acceptable error. In an emergency, a higher inclusion error is accepted in order to reduce exclusion, because excluding a vulnerable household can cost lives; in a long-term programme with limited resources, a better balance is sought. Reducing one error often increases the other, as well as the cost of targeting: targeting that costs more than the assistance distributed to wrongly included households makes no sense.

The verification process

1. Public reading of the provisional list, in the presence of men and women, with a reminder of the criteria.
2. Appeal period of 3 to 7 days, with several channels (help desk, telephone line, box, committee).
3. Visits to a sample of 10 to 20% of registered households by a team different from the one that drew up the list, and to all contested cases.
4. Correction and new reading; documentation of changes.
5. Duplicate checks against other organisations' lists, in line with data-sharing agreements.

8Worked example: choosing the modality and targeting in Haut-Kéla district

Let us return to the fictional Haut-Kéla district (see day 1). The cluster allocates two areas to "Terres Solidaires": the displacement sites in Doumbaré (around 3,500 displaced households) and the Sarakoro agropastoral zone (11,800 households, of which the assessment estimates that 35% will face a deficit during the lean season). Available budget: 2.75 billion currency units (CU) for six months, including direct implementation costs.

1. Data for the decision

2. Comparison matrix

Each option is scored from 1 (unfavourable) to 3 (favourable) on each criterion; the weighting is agreed by the team before scoring to avoid "engineering" the result.

Decision: mobile money transfers in Doumbaré, with an option to collect cash at an agent for people without a phone or who prefer it; in Sarakoro, a mixed modality: in-kind food prepositioned for the two months when the road is cut, and cash paid out in the commune capital in the other months, in a single monthly payment on dates not publicly announced in advance. The matrix does not replace judgement: it forces the trade-offs to be made explicit.

3. Comparative cost

For a monthly transfer worth 54,000 CU per household in Doumbaré: the cost of mobile cash (2% transaction fees, registration, SIM cards for households without one, staff) is estimated at 1.13 CU per CU transferred; for food procured regionally and transported (purchase, transport, storage, 1.5% losses, distribution), 1.38 CU per CU of value received, with the value calculated at local market prices. Over six months and 3,500 households, the difference amounts to around 280 million CU, the equivalent of five months of assistance for some 900 additional households.

4. Targeting in Sarakoro

1. Criteria, derived from the assessment and discussed in focus groups: fewer than 5 small ruminants and less than one hectare cultivated; or a household without an able-bodied adult; or a household with a child being treated for acute malnutrition; or a household headed by an isolated older person or a person with a disability. The last three categories are automatic inclusion criteria.
2. Targeting committees in the 23 villages: 7 members, including at least 3 women and a representative of older people or persons with disabilities, with no administrative or traditional authority as chair.
3. Drawing up the lists by the committees in the presence of the team; registration of households by phone with a photo of the recipient, and choice of recipient made together with the household (the person who manages food, most often a woman, unless there is an objection).
4. Verification: visits to 15% of registered households and 5% of non-registered households by a separate team, using a short form (assets, livestock, consumption the previous day). Result: inclusion error estimated at 14%, exclusion error at 21%. Two villages account for half of the inclusion errors: their lists are redone with a new committee.
5. Complaints: 312 complaints received during the appeal period, including 241 requests for inclusion; 96 households added after a visit.
6. Prioritisation: the final list contains 4,380 households whereas the budget covers 4,100. The steering committee decides to keep all households in the automatic categories and to reduce the duration of assistance to five months for the others, rather than reducing the transfer value below the calculated gap.

What the example illustrates: the modality varies between two areas of the same programme; the matrix makes trade-offs explicit; targeting errors are measured and corrected, not assumed; the budget constraint is managed through a transparent prioritisation rule, not through a silent reduction in transfer values.

9Protection, gender and accountability: building risks into the design

The Core Humanitarian Standard on Quality and Accountability (CHS, revised in 2024) sets out nine commitments, several of which directly concern the food response: assistance that is appropriate to needs, accessible and timely, that does no harm, that is based on people's participation, with safe and responsive complaints mechanisms. The IASC commitments on accountability to affected people and the IASC six core principles on protection from sexual exploitation and abuse apply at every stage, from targeting to distribution.

Risk analysis by stage

The complaints and feedback mechanism

• Multiple channels: toll-free hotline, help desks at the sites, boxes, committees, direct contact with staff. Channels are chosen after consulting women, older people and persons with disabilities.
• Categorisation: requests for information, requests for inclusion, complaints about quality, sensitive complaints (fraud, sexual exploitation and abuse, staff misconduct). Sensitive complaints follow a separate, confidential procedure, to the PSEA focal point or the investigation unit.
• Timeframes: immediate acknowledgement of receipt; response to non-sensitive complaints within 7 to 14 days; feedback loop closed with the complainant.
• Analysis: monthly trend report shared with the programme team, with the decisions taken. A mechanism that receives no complaints is not a sign of quality, but rather of inaccessibility or lack of trust.

Gender: beyond the female recipient

Systematically designating women as recipients can improve the use of assistance for food, but it can also increase their workload or expose them to violence. Good practice is to analyse, in separate focus group discussions, who decides on spending, who goes to the market and what risks each person faces, then to propose a default rule with the possibility of exceptions, and to monitor the effects in PDM (questions on disputes linked to assistance, on decision-making and on safety).

10Variations by context, common mistakes and donor expectations

Variations by context

Common mistakes

• Choosing the modality before the analysis, then writing the analysis to justify it.
• Confusing stated preference with feasibility: households may prefer cash while the market cannot absorb the demand.
• Letting local authorities draw up the lists on their own.
• Publishing vague criteria ("the most vulnerable") that cannot be verified or challenged.
• Not measuring exclusion, for lack of verification among non-registered households.
• Reducing transfer values to cover more households without saying so, which pushes everyone below the threshold at which an effect is achieved.
• Forgetting host communities, which creates tensions and exclusion.

What donors and auditors look for

• ECHO: justification of the modality based on market and risk analysis, comparison of cost-efficiency and cost-effectiveness, use of cash transfers where appropriate, and compliance with its guidance on large-scale cash transfers (segregation of functions, interoperability, controlled administrative costs).
• BHA: explicit justification of the modality and transfer value, mandatory market analysis, explanation if in-kind commodities are needed, market monitoring plan.
• Auditors: traceability of targeting (written criteria, committee minutes, signed lists, documented corrections), segregation of duties between the team that targets, the team that pays and the team that verifies, management of conflicts of interest among staff and committees.
• Evaluators: relevance of the modality in light of preferences and markets, measured targeting errors, consideration of at-risk groups.

11Interview questions and key references

Interview questions

1. "How do you decide between in-kind food and cash?"

A good answer cites the criteria (market functionality and elasticity, disaggregated preferences, protection and security risks, timeliness, cost-efficiency, capacity, legal framework, coordination), mentions at least one market analysis tool and stresses regular review and a fallback plan. It avoids any position of principle.

2. "Your budget covers 60% of eligible households. What do you do?"

The answer distinguishes prioritisation from targeting, proposes a transparent rule (severity, priority categories, areas), explains why reducing transfer values for everyone is generally a bad idea, mentions coordination in the cluster to fill the gap and evidence-based advocacy with donors, and communication with communities.

3. "How do you measure targeting errors?"

The answer describes verification with a sample of registered and non-registered households using an independent indicator, the calculation of inclusion and exclusion rates with their denominators, and the use of complaints. It acknowledges the trade-off between the two errors.

4. "A traditional leader insists on validating the list. How do you handle it?"

The answer recognises the leader's role and influence, involves them in information-sharing and mediation, but not in the final decision; it relies on a representative committee, public criteria, independent verification and the support of the coordination mechanism if needed. It mentions the risk to complainants and the protection measures.

5. "What does 'do no harm' mean in practice in a cash transfer distribution?"

The answer cites protection risk analysis, measures on the choice of recipient and payment security, data protection, prevention of sexual exploitation and abuse, monitoring of tensions within communities and families, and examples of actual adjustments.

Key references

• The Sphere Handbook (2018), food assistance and livelihoods standards; section on market-based programming.
• Minimum Standard for Market Analysis (MISMA) and response analysis tools: CaLP.
• RAM and MAG guidance: ICRC and International Federation of Red Cross and Red Crescent Societies.
• EMMA Toolkit: Oxfam, IRC, Practical Action.
• Operational guidance on targeting and prioritisation: WFP; joint guidance on targeting: UNHCR and WFP.
• Core Humanitarian Standard on Quality and Accountability (CHS, 2024 edition).
• IASC revised commitments on accountability to affected people and PSEA core principles.
• European Commission Communication on humanitarian food assistance (2010) and ECHO guidance note on large-scale cash transfers.
• Emergency Application Guidelines: USAID Bureau for Humanitarian Assistance (BHA).