Intrahousehold resource allocation : an overview

The policy failures associated with inappropriate acceptance of unitary models of household behavior are more serious than those associated with inappropriate acceptance of collective models, contend the authors. They support this claim with illustrations. Consider, for example, the effect of public transfers made to households. The unitary model predicts that the impact of such transfers is unaffected by the identity of the recipient because all household resources are pooled. With the collective model of the household, the welfare effects of a transfer may be quite different if the recipient is a man, say, rather than a woman. Most of their arguments for the policy relevance of model choice are based on the failing of the unitary model rather than on the strengths of a particular collective model. As a set, collective models may resolve some of the anomalies that have accrued under the unitary model, but further work is necessary to improve their predictive power. The authors admit to raising more questions than answers - which they regard as positive, considering that a conference in the late 1980s focused on whether it was even worthwhile going inside the"black box"of the household. The response to that question was that is was worthwhile examining household behavior, but few more definite answers have emerged, for three reasons. First, by their nature, the results of gender and intrahousehold analyses are specific to cultures and difficult to generalize, although the process of analysis can be generalized. Second, there is a lack of consensus about which conceptual model of the household to use, both across and within social science disciplines. And third, the collection of many intrahousehold data sets is not driven by policy questions. The challenge, the authors say, is to produce generalized results useful for policy formulation. In that regard, it seems desirable to apply a common conceptual approach to the analysis of policy-oriented case studies from a regionally diverse set of countries. Hypotheses about these studies could be developed and tested with and without the benefit of intrahousehold information to carefully measure the tradeoffs between the additional project and policy insights derived (and mistakes avoided) and the extra burdens of the analysis itself.