Agriculture & Future Food Systems · Article

Measure what assistance reaches and changes

Deliveries, actual use and remaining barriers are different observations. Monitoring needs clear denominators, follow-up and attention to effects beyond recipients.

Preparation and related reading

Shared Foundations

This article helps you

Separate processes, outputs and outcomes.
Interpret reach and use with explicit missing responses.
Look for unintended effects without assuming their cause.

A delivery is one step in an access response

A programme records that food reached a collection point. This is useful delivery evidence, but it does not show whether every intended household collected it, could prepare it or found it suitable. Following the response beyond the handover gives a more useful picture of access.

Monitoring means regularly collecting and interpreting information to guide decisions. WFP distinguishes processes, outputs and outcomes in its monitoring framework. A vehicle arriving is a process observation; a recorded handover is an output. Whether the support addressed the intended food-access problem needs further evidence.

Define who and what are counted

State the intended group, place and period. Count unique people or households when measuring reach; count transactions or deliveries when measuring service activity. A household receiving several deliveries is not several households.

Define the denominator before reporting a percentage. It could be the intended group, the recipients or those answering a follow-up question. These answer different questions. Record changes in the eligible group and avoid interpreting a changing denominator as improvement in delivery.

Keep follow-up gaps visible

In an invented programme, 100 intended households are identified and 80 receive assistance. Of those 80, 70 answer a follow-up question; 56 report that they used the food as intended and 14 report that they did not. The other ten recipients' use remains unknown.

Delivery reach is 80% of the intended group. Reported intended use is 80% among the 70 respondents. The identical percentages have different numerators and denominators. The diagram follows the recipient follow-up; it does not turn reports into a direct measurement of consumption or nutritional improvement.

Original hypothetical household follow-up for one period
Record	Households
Intended group	100
Received assistance	80
Answered follow-up	70
Reported intended use	56
Reported no intended use	14
Recipient follow-up unknown	10

Follow-up among 80 recipient households

Original hypothetical follow-up. No validated consumption, nutritional effect or causal programme impact is claimed.

Read the diagram as text

Of 80 recipients, 70 answer and 10 do not. Of the 70 answers, 56 report intended use and 14 do not. The missing ten remain unknown; these are reports, not direct intake measurements.

Parts

80 recipient households: A subset of the 100 intended households.
70 answered: Only these households provide a follow-up answer.
10 unknown: No follow-up answer; do not classify as non-use.
56 report intended use: 56/70= 80% of respondents.
14 report no intended use: Reasons and material destinations require investigation.

Connections

80 recipient households → 70 answered: answered
80 recipient households → 10 unknown: missing
70 answered → 56 report intended use: reported
70 answered → 14 report no intended use: reported

Ask what prevented use

Follow up on the reasons behind a recorded result. Food may require unavailable facilities, collection may be difficult or the offering may not fit needs. Not using it as intended is not automatically waste: it may have been shared, retained or used differently. Those routes require their own records.

Compare experiences across relevant groups without assuming that all members share a barrier. Offer suitable ways to give feedback, including for people who did not receive support. WFP notes that satisfaction figures can miss concerns when people are reluctant to criticise a programme.

Look beyond the intended result

A response can affect waiting time, travel, unpaid work and the use of local services. A market-based response also needs observation of prices and supply, including the experience of people not receiving assistance. An effect should be investigated, not assumed from the transfer form.

If prices rise after a programme begins, the timing is a clue rather than proof of causation. Harvest conditions, transport disruption or wider price movements may also matter. Gentilini's transfer primer discusses possible market interactions and the need for context-specific assessment; it does not justify a universal claim that cash causes inflation or food distribution depresses prices.

Use the record to make a bounded change

Combine delivery records, follow-up answers, participant explanations and direct observations where appropriate. Check whether a change addresses the identified barrier, then review again. A headline percentage is most useful when its missing information and definition remain attached.

Report what was measured, what was reported and what remains uncertain. Short observations do not establish long-term stability or prove a causal impact. Link the monitoring result to the next practical decision while keeping nutritional assessment, product safety and wider economic analysis within their appropriate subjects.

Worked examples

Two percentages with different meanings

Using the invented record, calculate delivery reach and reported intended use among follow-up respondents.

Delivery reach: 80 ÷ 100 × 100 = 80% of intended households.
Reported use among respondents: 56 ÷ 70 × 100 = 80%.
Recipients not answering: 80 − 70 = 10 households; their use is unknown.

Both results are 80%, but one measures delivery reach and the other a report among respondents. They cannot be substituted for each other.

A denominator can change the rate

Two periods each record 80 households receiving support. The first intended group is 100 households and the second 160. What are the reach percentages?

First period: 80 ÷ 100 × 100 = 80%.
Second period: 80 ÷ 160 × 100 = 50%.
The delivery count is unchanged, while the stated intended group grew.

Reach changes from 80% to 50%. This does not mean that fewer households received support; the denominator changed.

Check your understanding

Can the ten missing follow-up responses be entered as zero use?
Show answer and reasoning
No.
Missing responses are unknown, not evidence of non-use.
Does reported failure to use food as intended establish that it was discarded?
Show answer and reasoning
No.
Sharing, storage or another use must be investigated before classifying a material route.
Does a price rise after a cash programme begins prove that the programme caused it?
Show answer and reasoning
No.
Other influences and an appropriate comparison are needed for causal attribution.

Key ideas

Keep processes, deliveries and actual outcomes distinct.
Attach the intended group, denominator and missing data to every rate.
Use feedback and contextual evidence to investigate barriers and unintended effects.

References

Sources supporting the explanations and examples in this article.