8A.11 - Evaluating development-aid outcomes

Contrasting health and recovery outcomes

Aid works through local institutions, people and conditions, so a transferable technique does not guarantee the same outcome everywhere.

The 2014–2016 West African Ebola outbreak illustrates a relatively successful emergency response in interrupting transmission, alongside severe losses and shortcomings. Isolation and treatment reduce infectious contact; contact tracing identifies people needing monitoring; safe burials reduce exposure; trusted community communication encourages early reporting. External finance, staff, laboratories and logistics can expand these capacities, but local health workers and communities make them usable. WHO's 14 January 2016 announcement reported the end of the latest outbreak in Liberia while warning that flare-ups could occur. A milestone in transmission control was not proof of permanent elimination or complete recovery of health systems. WHO control mechanisms, WHO outbreak history, dated WHO announcement.

Haiti after the 2010 earthquake shows why relief and sustained recovery can have different records. Oxfam's 2011 progress report documented emergency assistance and movement toward longer-term services; short-term access to water was a real goal even while permanent services remained difficult. The subsequent cholera crisis also exposed profound problems. On 1 December 2016, the UN Secretary-General apologised for the UN's role in the epidemic and called for a better-funded response. Aid providers themselves can contribute to harm; delivery totals cannot settle questions of protection, accountability or trust. Oxfam account, UN statement.

The comparison is not 'Ebola succeeded and Haiti failed'. A specific intervention can meet one objective and miss another. Emergency control depends on coordinated action and trust; lasting development also requires stable finance, functioning services and local accountability. Compare equivalent goals and time horizons before ranking cases. Health gains also do not prove gains in freedom of expression or minority rights; those require separate evidence.

Attributing change to aid

The relationship between aid and development is difficult to isolate because aid is directed toward need and interacts with many other influences.

If recipients' life expectancy rises after a grant, the grant may have helped; improved sanitation, domestic spending, income growth or the end of conflict may also contribute. Conversely, a worsening indicator does not prove the aid was useless if it prevented an even larger deterioration. Need can itself cause greater aid flows, so a correlation between high aid and poor health can reflect donors responding to crises rather than aid causing poor health.

A stronger evaluation combines a taught mechanism with baseline and follow-up data, a credible comparison and evidence of implementation. Compare similar communities with different exposure where feasible, checking pre-existing differences and other changes. Randomisation can strengthen some programme evaluations but may be impractical or unethical; observational comparisons must acknowledge selection and uncertainty. Follow-up should include people who left the programme, or a survey may retain only the most successful participants.

Separate efficiency from effectiveness. A cheap programme may deliver many nets per dollar, while poor use makes it ineffective at preventing illness. A costly remote service can be valuable because it reaches an otherwise excluded population. Also inspect unintended effects, including local producers displaced by imported goods, diversion to elites and loss of voice when donors dominate decisions. A credible judgement specifies the outcome, group and period and explains how much of the change can reasonably be attributed to aid.

Quintiles, Gini and unequal gains

Average development gains can coexist with worsening inequality when richer groups receive a disproportionate share of income or consumption.