Targeting and modalities: who to help, how, and with how much

1Why target, and what poor targeting costs

With resources below needs, targeting decides who receives aid. It produces two errors. The inclusion error: households that do not meet the criteria receive aid (waste, loss of trust). The exclusion error: eligible households do not receive it (direct harm, often to the most marginalised). Reducing one generally increases the other: very strict criteria exclude vulnerable households; broad criteria include households that do not need it.

Calculation. Of 1,000 households assessed, 400 are truly eligible. The project selects 450, of which 330 are eligible. Inclusion error: 120 / 450 = 26.7% of the households assisted are not eligible. Exclusion error: (400 − 330) / 400 = 17.5% of eligible households are left out.

In humanitarian work, the exclusion error is the more serious one when survival is at stake. A targeting policy must state explicitly which of the two errors it is more willing to accept, and why.

Targeting also has a cost: surveys, verification, complaints, social tensions. When more than 70 to 80% of an area's population is in need, a general distribution (geographic targeting only) is often fairer and cheaper than household-by-household targeting.

2Targeting methods

Methods are often combined in several stages:

• Geographic: choose the most severe areas (week 1). Simple and hard to dispute, but imprecise within an area.
• Categorical: observable criteria (household headed by a single woman, person with a disability, child under two, recently displaced). Transparent, but these categories only partly match real vulnerability.
• Community-based: a representative committee identifies households according to publicly explained criteria. Uses local knowledge, but risks elite capture and the exclusion of minorities. It is controlled through random verification and an independent complaints mechanism.
• Vulnerability score (proxy means test): a household questionnaire produces a score calculated from variables correlated with poverty (assets, housing, composition). More objective, but costly, opaque to people and sensitive to false declarations.
• Self-targeting: the aid attracts only those who need it (cash for work at a below-market wage, less valued products). Cheap, but it can exclude people who cannot work.

The choice depends on four factors: the proportion of people in need, the availability of reliable data, social cohesion (risk of tensions if targeting appears unfair), and the time available. In all cases, the criteria must be public and understood by the community, and a complaints mechanism must allow an exclusion to be challenged.

3In-kind, vouchers or cash transfers

Cash transfers now account for about one fifth of international humanitarian assistance. Comparative data show that, for equivalent results, they often cost less to deliver than in-kind aid, let people choose according to their priorities and support local markets. They are not always the right answer.

The risks of cash can be managed: local inflation (price monitoring, transfer caps), insecurity (electronic payment, split distributions), diversion by a household member (choice of recipient based on gender analysis), sanctions and counter-terrorism (vetting of financial service providers). The idea that cash would be diverted to alcohol or tobacco is contradicted by many studies: spending on "temptation goods" does not increase significantly.

Multipurpose cash (MPCA) covers several basic needs at once; it is coordinated within the cash working group, which sets a common transfer value so that organisations do not compete with each other.

4Calculating a transfer value

A transfer value starts from the minimum expenditure basket (MEB): the monthly cost of the goods and services a household needs to cover its basic needs (food, water, hygiene, rent, energy, transport, communication). It is calculated for a typical household size, from prices recorded in markets.

The transfer value is not necessarily the whole MEB: what the household already covers itself (income, other aid) is deducted, and the programme's objective is taken into account. Basic formula: gap = MEB − average resources of the target household.

Example. Monthly MEB for a household of 6: 210 USD, of which 130 for food. Average monthly income of targeted households: 75 USD. Gap: 135 USD. The cash working group has set a common value of 80% of the gap to avoid creating dependency and to preserve own income: 0.8 × 135 = 108 USD per month.

Inflation. If the price of the basket increases by 12% in three months, the real value of a fixed transfer of 108 USD falls to 108 / 1.12 = 96 USD in initial purchasing power. Programmes therefore include an adjustment rule: for example, revision as soon as the MEB changes by more than 10%, based on monthly price monitoring.

Household size: a fixed amount per household disadvantages large households. Programmes often use brackets (1-3, 4-6, 7 people and more) with a cap, to remain simple to explain.

5Payment mechanisms and data protection

The delivery mechanism is chosen after the modality: hand-delivered cash, payment agents, mobile money transfer, prepaid card, bank account. The criteria: network coverage, access of women and older people to phones, distance to cash-out points, fees, agent liquidity, and identification requirements ("know your customer").

Electronic payment reduces the risk of theft and improves traceability, but creates data risks: beneficiary lists shared with an operator, authorities or other actors, which can expose people at risk (refugees, minorities). The ICRC Handbook on data protection in humanitarian action recommends: data minimisation, a defined purpose, written sharing agreements, a data protection impact assessment before any new system, and a limited retention period.

Reconciliation compares three lists every month: registered households, payments sent, payments actually withdrawn. Discrepancies (uncollected payments, duplicates, phone numbers shared by several beneficiaries) are signs of error or fraud to be dealt with quickly.

6Post-distribution monitoring

Post-distribution monitoring (PDM) checks, two to six weeks after a distribution or transfer, that the aid arrived, that it is being used and that it is producing the expected effects. It covers a random sample of beneficiaries and has four blocks:

• Process: amount or quantities received, waiting time, distance, travel cost, safety, respectful treatment.
• Use: expenses made with the transfer, resale of in-kind goods, sharing.
• Outcomes: food consumption score, coping strategies, debt.
• Accountability: knowledge of the criteria and the complaints mechanism, satisfaction, problems reported (requests for payment, harassment).

A PDM is valuable only if it triggers decisions. For example: if more than 10% of beneficiaries report having paid something to be registered, an investigation is opened and the registration process is changed; if 40% resell the oil received, the composition of the basket is adjusted or the programme switches to cash.