Contents - Previous - Next

This is the old United Nations University website. Visit the new site at

Synthesis of concepts and research needs

Definition of well-functioning families or family social health
Importance of family social health in producing child growth and development
Family management
Caring capacity
Beliefs, rules, and goals
Family boundary maintenance
Proposed research model for measuring family social health
The need for further research


In chapters 1-8 we carried out the first two objectives of this document, to review the literature on research pertinent to the family in five disciplines and to formulate and test an integrated empirical model of family characteristics that determine child welfare and development. In this chapter we use the findings from the literature and our own analyses to propose and support a definition of family social wellness. We propose models for testing and using this definition for the design and evaluation of policies and programmes. Finally, we discuss research approaches and tools needed to support a family social health approach to development assistance.

Definition of well-functioning families or family social health

Viewed together, the family literature provides consistent criteria for defining family social health, wellness, or well-functioning that could be subdivided in a variety of ways. For the purposes of providing assistance to socio-economic development, we choose to group them under four dimensions that roughly correspond to four different domains of policy or intervention approaches. These dimensions also show an approximate congruence with the four spheres of action designated by the Family in Development initiative of the Agency for International Development (AID): the family as an economic system, the family as a social and biological reproductive system, political choices and the family, and families at risk (Agency for International Development 1991).

1. Family management, including skills in accessing and managing resources needed to sustain the family under normal conditions and during crisis (food, money, shelter, transportation, health care, education, etc.), and skills in governance of the family as a cooperative unit (decision-making, bargaining, problem solving).

2. Family caring capacity, characterized by sensitive and loving transactions between family members and by adequate technical knowledge and caring skills in physical maintenance, nutrition, health, socialization, and education.

3. Family beliefs, rules, and goals including explicit and implicit ideals and values for the family itself, goals for individual members, resource-sharing rules, cultural rules and codes for family behaviour, perceptual frameworks, and cognitive interpretations.

4. Family boundary maintenance, or structural integrity, covering the formation and dissolution of reproductive partnerships, birth control, child custody and fostering, the launching of adult children, care arrangements for the elderly, temporary separation, and death.

These dimensions are embedded in the family's interactions with its ecological context, and must also be considered from the perspective of family stage in the life cycle (Walker and Crocker 1988). The international development assistance field should find the contextual considerations familiar, as far more attention already has been given to the contextual community level than to families themselves.

Importance of family social health in producing child growth and development

We used our data on families in Indonesia and Nigeria to create the empirical models (see ch. 8) that illustrate the importance of social health at the family level. These models support the above definition, although the studies were not designed to gather data explicitly on each dimension. Our data sets included many covariables of malnutrition and mental development (demographic, social, economic, attitudinal, and behavioural) and these covariables showed expected modest correlations with growth and cognitive test scores. We found that models using these variables to measure the quality of family functioning (social health) and to explore the effects of family social health on the children's overall development showed stronger and more meaningful associations than the one-on-one relationships. We interpret these findings as an indication that family-level interventions that improve the child's total condition are likely to be more effective than vertical programmes that influence only one aspect at a time.

Family management

Family management enables the family as a unit to acquire and manage resources consistent with its goals and cultural codes. We call this "management," rather than "coping," because psychologists use the term coping to refer to only one aspect of management, the response to crisis. The management dimension of family social health corresponds to the AID sphere of the family as an economic system. It includes McMaster's "basic task area" (see Epstein, Bishop, and Baldwin 1984): earning a living, day-to-day home economics, planning, and seeking to take advantage of social services and new business opportunities. Proactive, upwardly mobile management styles are associated with favourable child outcomes. Although our model data were not collected to test the family management dimension, a number of our academic and health care measures reflect its presence.

Caring capacity

This dimension, which corresponds to the family as a social and biological reproductive system, has a loving and fun-loving component and a micro-level competency component: these include knowing how to breast-feed, play peekaboo, soothe a crying child, be sensitively verbally responsive, tell a story, prepare oral rehydration therapy, and clean a skin puncture; knowing when to go to the doctor; knowing how to protect food and kitchen utensils from animals. It captures the capacity to enhance and celebrate living through microlevel transactions between domestic partners, their children, and others.

The affective aspect of caring capacity involves management at the micro-level, where the behavioural transactions between individuals cross the perceptual boundary from "doing" to "being," as in "being" empathetic, warm, and sensitively responsive. The technical aspect includes child stimulation, nutrition, health care, sanitation, family planning, and other teachable skills - the common topics of parent education programmes.

Beliefs, rules, and goals

The beliefs, rules, and goals of the socially healthy family provide a support structure for economic success for both parents and children. They facilitate resource sharing and forward planning, which benefit all members. They are rooted in society's concepts of the ideal family, in cultural value systems that draw support from the cultural contexts of the community and society. Concerns for these value systems at the societal level correspond to AID's sphere of political choices and the family.

These values and ideals have an altruistic and an artistic component, reflected, for example, in the Yoruba proverbs about parenting (ch. 7), in pictures of the Madonna and Child, as well as in pictures by Norman Rockwell. Cultural ideals for the family give meaning to life. The ability to fulfil them is a criterion for judging the success of life itself. Cultural periods remembered as golden ages have been defined by the quality of family life.

Importance of cultural paradigms, legal and social entitlements, and cognitive interpretations

Research literature across the social sciences shows that important determinants of family social health include cultural and religious ideals and values, individual beliefs, and the "fit" between these paradigms and the social conditions required for economic prosperity. These paradigms are expressed in family law, social welfare entitlement, family size limitation, parenting codes, cognitive interpretation of events, and in gender and age-specific conventions for intra-household resource and task allocation. Examples of evidence from the literature include:

1. Sociology: continuing dialectic between socio-economic conditions, values, laws, and lifestyle prescriptions; negative effect claimed for values of capitalist economy on social responsibility; positive effects of US medical establishment's diet and exercise prescriptions on mortality indicators; dialectic process illustrated by current US debate over abortion.

2. Anthropology: many examples of resource allocation rules that negatively and positively affect categories of individuals. Observation that children in certain modernized cultures are less responsible and altruistic than children in more traditional societies.

3. Household economics: recognition that implicit cultural contracts and legal and institutional factors that determine the disposition of household assets, such as family laws regarding property rights and social entitlements, play a greater role in intra-household resource allocation than active bargaining among family members ("shoot-out at the family corral").

4. Psychology: the extreme resistance to change of male gender roles in dualearner households in the United States where the wife earns as much as the husband but continues to perform most domestic tasks; the finding that the main determinant of why some families cope while others fall into crisis under similar stress is the meaning that the stressful event holds for the family and the individuals within it.

5. Development assistance: international assistance agencies historically have been careful to remain neutral in their approach to values in order to build consensus across cultures. Ideologies put forth by these agencies, such as Child Survival, have played a major role in shaping programmes and effecting evaluation indicators. With the worldwide movement towards deepening democratic reform, however, AID (1991) has recognized that cultural adjustment of legal and social structures is needed to reduce intrafamily inequities.

The cultural renewal process: How paradigms change

The renewal process by which cultural value systems adapt to support socio-economic improvement also can be considered as a form of "cultural adjustment" that is needed in parallel to structural adjust ment. Historically, cultural paradigms of the family change from within through a dialectic, self-corrective process. Opinion leaders within the culture formulate and voice these changes. Ideologies co alesce around political and economic forces and the mood of the times; they both respond to and are precognizant of socio-economic change. Certain orientations have deep cultural roots, as for example the 3,000-year-old Jewish tradition. Other social trends appear as chaotic as the weather. Cultural consensus continually reshapes from within.

Positive change in culture-wide codes, such as healthful diet changes in the United States, appear to occur when the local intelligentsia first develops and articulates the new codes through consensus building in professional circles and then through the media, social services, and legislation, with ongoing feedback via professional meetings and journals and mass communications.

Studies of changing parenting codes also reveal a self-regulating, adaptive process when underlying resources are sufficient to permit families to regenerate. The greater the cultural distance between the traditional and modern lifestyles, the greater the paradigm shift required. In many of the world's poorest countries, cultural renewal has not kept pace with social change because of the dearth of resources available to the intelligentsia, which must formulate and legitimize new codes. Such moral rearmament activities are trivialized and ring hollow when they are left solely to the propaganda wing of the government in power. Cultural reconstruction becomes deadlocked by conflict between social profitability for the culture as a whole and the survival needs of its individual families. Values supporting small family size, for example, cannot prevail as long as individual families cannot afford to limit the number of their children. Cultural renewal also is deadlocked by resistance to necessary reductions or downward adjustments of entitlement to groups who continue to hold power.

In the most seriously affected parts of the world, moral codes developed over millennia of social evolution have disappeared with the lifestyles they supported. The absence of adequate cultural renegotiation and regeneration by legitimate opinion leaders has produced social conditions resembling cultural wastelands, over which unprincipled profiteers, degenerate gangs, and religious fundamentalists vie for control. Chapter 10 suggests actions that may assist developing countries in cultural renewal.

Family boundary maintenance

Flexible boundary maintenance, adapting to changing life-cycle needs, is a characteristic of well-functioning family systems. Involuntary boundary problems occur during armed conflict, sudden death, migration to find paid employment, and extreme poverty that forces reproductive partners to abandon each other and their children. Pathological boundary crises include male abandonment and separation leading to female-headed households; extramarital or conflicted polygamous relationships; struggles over child custody, support, and fostering; contested inheritance and property rights; and predation by some family members on others. These pathologies often involve problems with boundaries between family members, reflected in violence, sexual abuse, and dysfunction involving substance abuse. Families classified "at risk" typically have boundary problems.

Proposed research model for measuring family social health

Instrument development for the evaluation of family support and parent education programmes is still in its infancy. Krauss (1988) reviewed more than a dozen measurement instruments that have attempted to assess family functioning. Use of these tools has resulted in inconclusive outcomes and generally has failed to establish hard links between programmes, family measures, and individual wellbeing. Nevertheless, these instruments provide a smorgasbord of concepts and items to be considered in future research.

The use of a family social health approach in programme design and evaluation calls for new modelling procedures that assess the following:

1. How family social health determines the demand for inputs provided by programmes;

2. The impact of those inputs on the family as a whole and on individual family members;

3. The effects of family social health on child outcomes under conditions of changing inputs.

This formulation draws on ideas expressed by Berman, Kendall, and Bhattacharyya (1994). The best ways of accomplishing this task require further discussion. Dunst and Trivette (1988), for example, use hierarchical regressions and canonical correlations for family systems analysis. We suggest the development of structural latent variable models similar to those presented in chapter 8, but which incorporate certain desirable features of regressions. The coefficients in the models in chapter 8 cannot be used to estimate effect sizes. These models were not designed to incorporate policy variables (e.g. government expenditure on education and other programmes created outside the family). The hybrid models that we propose would permit prediction of the impact of different levels of programme resources on targeted outcomes, but would continue to take into account the integrity and complexity of the family system. Such models would incorporate both supply and demand relationships. The UNICEF conceptual framework might serve as a starting point for conceptualizing analyses that test the effects of multiple existing vertical programmes on households and families.

The need for further research

The boundaries between the methods and orientations of the different social sciences increasingly overlap. This merging of disciplines with a redefinition of paradigms has been claimed to be a part of the trend towards "dedifferentiation" in the post-modern era (Lash 1990). For policy-related work, disciplinary interests recede in importance as the topics of study spring into the foreground. The family is a topic that demands this blending and synthesis of methods. In fact, the family could prove to be the best single topic for interdisciplinary methodological development, because within-disciplinary methods for family studies already are well articulated and explored, providing a tested menu of methods and results to contrast and compare.

Fields that are multidisciplinary in origin, such as international nutrition and food policy, may be at the forefront of disciplinary pluralism. The same nutrition-related theses are produced by academic departments of nutrition, epidemiology, international health, psychology, human development, human biology, sociology, anthropology, economics, and demography. Social scientists working in these already multidisciplinary fields may be more experienced and objective in bringing a cross-disciplinary perspective to the family than researchers based in disciplines with long-established family studies.

New greatly increased computing capacity invites statistical analysis plans that draw eclectically from the various disciplines and synthesize elements from them into increasingly comprehensive explanatory models. For example, endogeneity - bidirectional flow of causality previously modelled by econometricians - now can be taken into account in epidemiological models using packaged programs for simultaneous logistic regression. Conversely, the structural equation models that we find in psychological research can be adapted along econometric lines to permit cost-effectiveness estimates for varying development inputs, as we suggest above.

Antecedents of new, more inclusive modelling may be drawn from systems modelling in development studies that began in the early 1970s. The international nutrition planning movement provides a relevant example. In the second half of the 1970s, multidisciplinary research teams in institutions in both industrialized and developing countries designed elaborate cross-disciplinary conceptual frameworks. In the late 1970s and early 1980s, disenchantment with the mathematical modelling capabilities of the time, and failure to couple systems theories to action, swung the pendulum towards less formal and less quantitative approaches to operational research and evaluation, in which the systems relationships served as conceptual frameworks.

As the moment approaches when numerical methods will be able to keep pace with programmatic action, the attempt to quantify cross-disciplinary theory building re-emerges as a central priority. The current and coming generations of computers provide sufficient computing power for a quantum leap in statistical social science applications. These applications have the potential for creating portable computer field tools that remove many of the time lags and much of the guesswork from global monitoring and development assistance. Many of the component pieces and features that should be incorporated into these applications already are evident. What has fallen furthest behind is statistical programming. A new, well-publicized crossdisciplinary statistical programming agenda for the social sciences is critically needed to stimulate the creation of the needed software packages.

This call for theoretical and statistical research and development is applications driven. Computers of the new generation already are in our homes and overseas field offices, where their potential is underused. The payoffs for such development work should be comparable to the increments in efficiency gained by moving from boat mail to airmail to fax to e-mail. We suggest that the following should be done:

1. Develop an intradisciplinary research team on the family to:

(a) Evaluate cross-disciplinary findings and formulate new theories based on this evaluation.

(b) Determine research agendas and data analysis needs.

2. Hold a meeting of statisticians serving these disciplines to define statistical research and computer software development to meet the needs of the research agenda. Work will be needed in the following areas:

(a) Non-parametric, non-linear, and robust versions of all multivariate procedures, such as partial correlation, multiple regression, ANCOVA, MANOVA, to deal with data not nor mally distributed, or non-linear, and that may have problematic outliers. At present, statistical software for univariate analysis copes with these issues whereas software for multivariate analysis does not, with very few exceptions.

(b) Easy-to-use jackknifing procedures that permit accurate estimation from smaller samples (essential for field procedures too burdensome to conduct on very large samples).

(c) Flexible, multivariate repeat-measures programs - e.g. MANCOVA packages that permit individualized specification of needed covariate and other calculations.

(d) Better methods for dealing with statistical significance when conducting multiple tests on large sets of variables - e.g. more programs for specifying combined variable models for testing subsets of coefficients together (methods that do not demand data normality).

(e) Methods or rules for making multivariate models more robust in that the coefficients they produce need to become less sensitive to the entry or removal of individual variables. In the present state of the art in model building, the almost arbitrary addition or removal of individual variables from a large data set can create almost any outcome desired by the data analyst.

(f) Better programs for sequence recognition and for analysis of interactive systems.

(g) Better software for qualitative data analysis in formats that can be merged with quantitative data sets.

(h) Management information systems (MIS) applications that conduct more automatic data quality checks and that have built-in multivariate analysis programs, customized to the MIS data and robust against the effects of outliers.


Agency for International Development (AID). 1991. "Family and Development Progress Update, April 1991."

Berman, P., C. Kendall, and K. Bhattacharyya. 1994. "The Household Production of Health: Integrating Social Science Perspectives on Micro-level Health Determinants." Social Science and Medicine 38: 205-215.

Dunst, C.J., and C.M. Trivette. 1988. "Toward Experimental Evaluation of the Family, Infant and Preschool Program," In: H.B. Weiss and F.H. Jacobs, eds. Evaluating Family Programs. New York: Aldine de Gruyter, pp. 315-346.

Epstein, N.B., D.S. Bishop, and L.M. Baldwin. 1984. "McMaster Model of Family Functioning." In: D.H. Olson and P.M. Miller, eds. Family Studies Review Year-book, Volume 2. New Delhi: Sage Publications.

Krauss, M.W. 1988. "Measures of Stress and Coping in Families." In: H.B. Weiss, and F.H. Jacobs, eds. Evaluating Family Programs. New York: Aldine de Gruyter, pp. 177-194.

Lash, S. 1990. "Modernity or Modernism? Weber and Contemporary Social Theory." In: Sociology of Postmodernism. London: Routledge, pp. 130-133.

Walker, D.K., and R.W. Crocker. 1988. "Measuring Family Systems Outcomes." In: H.B. Weiss and F.H. Jacobs, eds. Evaluating Family Programs. New York: Aldine de Gruyter, pp. 153-176.

Contents - Previous - Next