The following principles, or general guidelines, are recommended for consideration when designing an early childhood care and development program. They derive in part from the nature of early childhood care and development (as described above), in part from social and economic requirements for programming in the IDB, and in part from the experience of others in designing effective and sustainable programs.

• While trying to reach as many children as possible, give priority to children and families living in conditions that place the children at risk of faltering in their development.
• Place programs in a larger policy and programming framework promoting a multisectoral and integrated (or at least convergent) strategy based on a holistic view of child care and development. Look for ways to incorporate an ECCD component into ongoing programs.
• Be open to diversity and to complementary approaches that focus on families and communities as well as children and that respond in different ways to the needs of children from conception until entry into school.
• Adopt a participatory and community-based approach in both the process of design and the program.
• Support and build on local ways of coping with problems of child care and development (that is, take a constructive rather than a compensatory approach).
• Favor programs that take an intergenerational view, meeting the needs of adults (especially mothers) as well as children.
• Adjust programs to the developmental stages of the children in them and allow for differences among children.
• Seek quality, taking into account appropriate practice as defined in relation to how children develop and in the cultural context, and relate quality to results.
• Seek cost-effectiveness and financial feasibility.
• Incorporate monitoring and evaluation into programs from the outset.

There are some obvious tensions and potential tradeoffs among some of these guidelines. A program design simultaneously targeting scale, quality, comprehensiveness, adjustment to local conditions, participation, cost-effectiveness, and financial feasibility is bound to demand choices. But there are ways of reducing or working around such tensions. In large-scale programs, for example, there is a tendency for quality to be watered down and for local participation to disappear. But if scale is defined in terms of the sum of smaller programs and if the criterion for measuring scale is not the entire population but those most at risk, the tension between scale and quality and between scale and local participation can be greatly moderated. (See tab 4 for a discussion of several of these cross-cutting issues.)

If a program is to be large scale, good quality, and comprehensive, costs may get out of hand, requiring compromises. The tendency in such situations is to ask which program components can be cut or which will provide the most results. Another way to compromise is to retain a comprehensive approach in the name of quality and effectiveness while reducing scale by focusing services on populations at risk, identified using more rigorous criteria. Another way is to phase in components, using a local planning process to determine which component should provide the starting point for action in the community.

These decisions will also involve tension among political, social, economic, and technical criteria. The predominant political criteria are the program's size and control. Size signals activity and presence and can help ensure political support; quality and effectiveness, as reflected in outcomes, are less important as political criteria. Control is usually exercised from the center (even in so-called decentralized programs).

The predominant social criteria are equality (bridging gaps, improving distribution), respect for the culture, and participation. A social focus suggests targeting groups at risk, constructive rather than compensatory programs, and participatory planning. But permitting participation that empowers the community may be anathema to the dominant political forces because it implies ceding control. So the political decision may be to avoid real participation.

The predominant economic criterion is cost-effectiveness. This criterion can lead to the choice of a model that is considered most effective, regardless of the population it serves best. Or it might be applied within a social framework to the programs judged to be most equalizing and empowering. The economic criterion may trade off against a political criterion of reaching the most people possible.

The predominant technical criteria are quality and improved outcomes. Here, choices may be influenced by a desire to apply the latest technology. But using the latest technology does not necessarily best serve children at risk or even improve outcomes. Moreover, improvements in outcomes cannot be sought without considering cost.