In some cases this decision will appear relatively easy. It may already have been made by those requesting Bank funding. Or it may not need to be made because the ECCD project is part of a broader project (in health or urban development, for example) for which the population has already been defined. Or a national plan may provide specific guidelines for directing social projects. In many cases, however, the decision is not at all clear, or it needs to be debated, or it must be made much more specific.

Choosing the focus—children or others?
It is logical to assume that the main beneficiaries of a project relating to early childhood care and development will be children. But potential beneficiaries of a project focused on children and their welfare will include families, communities, and even the society at large. And even if the main beneficiaries of an ECCD project are children, the main participants or groups to whom the project is directed may not be children. Instead, the project may focus on one or more of the following groups:

• The population at large—if the project seeks to disseminate information about child rearing practices to caregivers and citizens through the mass media or as part of a broad education program.

• Politicians, lawmakers, and planners—if the project's goal is to change or create policies and laws in favor of early childhood care and development.

• Teachers, health care workers, or NGO or government staff—if the project's goal is to improve social institutions responsible for children's programs, including through the training or retraining of staff.

• Community leaders and residents—if the project is concerned with community organization and action in support of child development.

• Family members or others responsible for immediate caregiving—if the project seeks to improve child care practices.

In all these cases the emphasis is on raising the consciousness of or educating adults who will, through their actions, influence the development of young children, sometimes directly, sometimes indirectly. Ideally, an ECCD program would include strategies focused on all these groups. But that rarely proves possible in the real world. Choices must be made. And the populations identified will need to be reached in different ways and with different models.

Deciding on social and economic characteristics
Projects aimed at improving the welfare of children sit within a broader social context. All projects receiving support from the IDB are expected to bear directly on its twin objectives: reducing poverty and promoting social equity. Among the first criteria for delineating the population to receive preference in a project, therefore, is a social criterion.

Tab 2 suggests a range of social indicators that might serve as the basis for delimiting the general population of children and families to be included in a project. A key criterion might be poverty, defined by income level or by lack of access to services. Or the project population might be defined using a set of risk factors that make some children more vulnerable than others to delayed or debilitated development. These risk factors could include family structure (single-parent families or female-headed households), employment of parents in the informal sector, or relatively low education levels of parents and other family members.

The indicators of risk, poverty, or discrimination chosen to delineate the project population should be closely related to indicators of children's developmental status. If they are not, they probably are not good criteria to apply in an ECCD project.

Setting age limits
As tab 1 indicated, early childhood runs from conception to about age six, and it makes sense to extend the period into the first year or two of primary school in order to deal with that period of adjustment. It was also suggested that the period can be broken down into a series of shorter periods corresponding roughly to maturational changes in children that determine particular needs and open up new possibilities for learning. Thus a project must either focus on a relatively narrow age group and its corresponding stage of development, or be prepared to respond in different ways to children of different ages and developmental stages. For example, an ECCD project might focus on development during pregnancy, a child's first year, the period from about age one to three or that from age four to six, the preschool year, or some combination of these.

As the next sections show, the age of the children on whom a project will focus has important implications for the project's design. If a project focuses on children in their earliest years (say, from birth to age two or three), for example, it is likely that the project will be directed more immediately to adults rather than to children, that it will emphasize physical development, that the counterpart organization will be in the health sector rather than education, and that the organization will involve relatively small ratios of adults to children.

Responding to developmental status and need
It is reasonable to presume that poverty and discrimination are closely related to children's condition and therefore to their need. But there are at least two reasons for considering developmental status and need as a separate criterion for identifying children and families. First, some children who are relatively well-off economically or are part of a dominant social group may nevertheless be very much in need of assistance. Being better off economically and socially does not guarantee that a child will be treated well by the family—abuse and lack of affection are not confined to a particular economic or social group. So a decision must be made about whether a project should include children whose social condition cannot be termed disadvantaged.

Second, some children who fall in the at-risk category will nevertheless thrive, even without project assistance. Their families may have sound child rearing beliefs and practices, or the family structure or social support system may provide functional caregiving options. Should a project include such children and families?

There is another, practical reason for using developmental condition and need as a defining characteristic of the project population: using only social criteria and age groups to define a project population may result in an unmanageable number of potential beneficiaries and participants. It would probably be impossible for an ECCD project to reach all families considered at risk, for example, if poverty is the criterion and half of all families in the country live in poverty. One way to focus the population is to make the social criteria more limiting or to define the age group more narrowly. Another is to examine children's developmental status and direct the project to those most in need within the broader social and age categories. That implies a need for information about the developmental status of particular children

But as tab 2 reported, few countries in the region have a system in place for periodic measurement of children's developmental status, defined as including mental, social, and emotional as well as physical development. More common is sporadic use of a variety of instruments purporting to measure child development. Where an instrument for measuring development is used, it can be drawn on, but with awareness of the purposes in applying the instrument and its reliability, validity, and cultural appropriateness. Local experts could assist in interpreting the instruments and their results. Instruments in use in the region for measuring children's developmental status include the following:

• Brazil: Escala de Desarrollo de Heloise Marhino (EDHM).

• Chile: Escala de Evaluación de Desarrollo Psicomotor (EEDP); Test de Evolución Psicomotor (TEPSI).

• Colombia: Pautas de Desarrollo Infantil en los Primeros Seis Año; Escala Abreviada de Desarrollo (adapted for use in Bolivia); Escala de Valoración Cualitativa del Desarrollo Infantil.

• Costa Rica: Escala de Desarrollo Integral del Niño (EDIN) (adapted for use in Guatemala).

• Jamaica: child status profile.

• Mexico: Escala de Desarrollo; Hoja Gráfica de Desarrollo.

The Denver Developmental Screening Test also is widely used in the region. Many countries include in their health carnet indicators of the physical development of children. Proxies of development can also be used, such as the nutritional and health status of children based on standard indicators like weight or height for age. But program developers should keep in mind that the correspondence between nutritional status and mental or emotional development is far from perfect. For more detailed information about the instruments listed above and about others for determining risk factors in the environment see references listed in the section on measurement in tab 8.

The main point here is that an effort should be made to establish the actual condition of children's development and instruments are available to help do that. Describing the developmental status of children in a specific setting can help define the population of children to whom the project will be directed. It can also help set a basis for evaluating the project later on.

Stakeholders probably will have their own sets of criteria for determining which communities or families or children should be included in a project-or which should be excluded. For this reason it is important when talking with stakeholders to have as clear an idea as possible about the social condition and needs of families and children according to "objective" criteria.

Choosing between universal and partial coverage
Sometimes the first question asked when selecting a project population is, Should the project seek universal coverage or be directed to a population defined by poverty or by social discrimination? An argument is sometimes made that coverage should be universal, to "democratize" a service. Universal coverage of preschool programs for the year immediately preceding primary school is now being sought in some countries, for example.

But arguments for universal coverage of any project must be treated cautiously (see tab 4). Although appealing for both political and social reasons, universal coverage defined numerically may not be universal in its benefits to children. The quality of an ECCD service may not be the same everywhere, only in part because universal treatment makes it difficult to adjust the service to local and cultural conditions. So a universal strategy, such as universal preschooling during the year before primary school, will not necessarily have the desired effects on poverty or social discrimination.