Unfortunately there is usually a gap between good intentions and effective poverty relief interventions. People who are trapped in poverty are often there because they live in a system which operates to keep them poor. The high costs of insecurity, lawlessness, and exploitation may outweigh any savings that a person might otherwise have made, so that the person remains poor no matter how hard he or she works.
As a result of experiencing the failure of direct poverty interventions, most donors seek to influence the outcome by requiring their gifts to be used in a particular way, such as in skills training, or microfinance loans. However, most donors do not follow up their donations sufficiently to be sure about the degree to which their interventions have had a positive outcome. Usually the outcome is far less positive than expected.
An organisation called GiveDirectly works on the assumption that, although the simplest form of assistance, giving cash, has many known drawbacks, it is actually more helpful and efficient than most of the other types of intervention. GiveDirectly monitors its results using randomised control trials, and its model presents a challenge to other types of poverty relief: if an intervention is not at least as successful as giving cash, how can it be justified?