Assessing outcomes: benefits, failures and harms
Non-governmental organizations (NGOs)
Evaluate a non-governmental organization through measured results, safeguarding and accountability rather than its label.
Evidence reviewed: 17 October 2026. Roles, financial periods, allegations and legal outcomes retain their stated dates. Later appointments are distinguished from officials already serving.
Begin with a specific activity
A non-governmental organization (NGO) can deliver services, distribute resources, investigate abuses or campaign for policy change. Those activities require different measures. Counting meetings or money distributed describes outputs; establishing whether people became safer or better off concerns outcomes. Neither an admirable mission nor a large budget settles that second question.
A measured benefit
An experimental study by Dennis Egger and colleagues examined GiveDirectly cash transfers in rural Kenya. It found increased consumption and assets among recipients, positive effects on nearby households and businesses, and limited price changes in that setting. This is evidence about a particular intervention and population, not a guarantee that every cash program or organization will have the same effects. [1]
A documented failure
The Charity Commission’s 2019 inquiry into Oxfam Great Britain examined safeguarding and governance, including its handling of misconduct in Haiti. It found serious failings. This is a regulator’s investigation of a named organization and period; it should neither be hidden behind the charity’s mission nor generalized into a finding that all aid organizations behave similarly. [2]
Correction also needs evidence
The regulator’s follow-up tested corrective changes under statutory oversight. An action plan is a starting point, not proof that every problem was solved. Keep original findings visible while checking implementation and remaining risks. [3]
A practical evaluation
Ask who benefited, who was overlooked, who might have been harmed and how a complaint could be raised safely. Compare observed changes with a plausible alternative rather than crediting the organization for everything that happened afterward. Consider the time period, independent checks and what the evidence cannot show. Potential harms such as exclusion or unsafe delivery should guide investigation; they are not findings against an organization until supported by evidence.
People, institutions and evidence
GiveDirectly cash-transfer intervention in Kenya
Relationship: Evaluated program
Documented record: A randomized study measured economic outcomes for recipients and surrounding communities.
Limits: The findings concern this intervention and setting, not all non-governmental organizations. [1]
Oxfam Great Britain
Relationship: Safeguarding investigation
Documented record: The regulator published adverse findings in 2019 concerning a defined organizational record.
Limits: Findings must retain their date and scope; they are not a sector-wide accusation. [2]
Oxfam Great Britain and Charity Commission
Relationship: Corrective supervision
Documented record: The follow-up describes an action plan, progress reporting and regulatory oversight.
Limits: A corrective process must be evaluated separately from the original failure. [3]
Related learning
England: charities, campaigning and separate legal entities
French collectives and court review: the 2023 dissolution case
Oxfam: safeguarding failures, reforms, and humanitarian work
Médecins Sans Frontières: medical operations and distributed leadership
External Resources
[1] Egger and colleagues: experimental cash-transfer results from Kenya
original research
[2] Charity Commission: 2019 inquiry into Oxfam Great Britain
primary regulatory inquiry
[3] Charity Commission: Oxfam safeguarding follow-up
primary regulatory follow-up