2. Plan small-scale, flexible projects. A plan should be a blueprint, not a prison. It should be able to incorporate new information that emerges during the project.

3. Let the people benefiting from the project make the decisions. The experts' job is to share their knowledge, not impose it.

4. Look for solutions that can be duplicated by hundreds and thousands for the greatest impact on development But the solutions must still be tailored to fit local needs.

5. Provide education and training, particularly for young people and women, who remain the most effective agents of change because they are bound to the realities of the family's survival.

6. Keep external inputs to a minimum to reduce dependency and increase stability Subsidies, supplements and inappropriate technology are unsustainable.

7. Build on what people are doing right. New ideas will be adopted only if they do not run contrary to local practice. New technologies must support existing ones, not replace them.

8. Assess impacts of proposed changes. A multi-disciplinary team, ideally including specialists from the same culture, should look at economic, social, cultural and environmental aspects.

9. Consider both inputs and outcomes. The failure of projects focusing on a single outcome, such as agricultural productivity, has proved that more is not always better.

10. Maintain or improve the participant's standard of living. Long-term improvements are unsustainable unless they address problems the poor face today.

Source: Rural Technology 8 (2), September 1991