particular. Moreover, most employed respondents firmly maintain that their children would be better taken care of if that responsibility did not have to be entrusted to others. Among these respondents who work only out of necessity, some recognize positive aspects of their employment, such as a sense of achievement, independence, and the opportunity it affords them to break out of the isolation of their homes. The effects of paid employment on gender relations within the household, however, do not appear to be uniformly positive. In the extreme situation when the respondent becomes the sole support of the household, the treatment she receives or puts up with seems to be worse than it otherwise might be.
As the various chapters included in this volume indicate, the pace and scope of the rise in contraceptive practice and the ensuing decline in fertility have been massive in the last two decades in most of Latin America. Child mortality, domestic arrangements, and female labour-force participation have also undergone quite dramatic transformation in many countries of the region. Yet, fully aware of the poor economic performance of the last decade in Latin America, most observers would be unwilling to believe that women and children in the region are better off now than they were ten years ago. It is a daunting setting in which to attempt an evaluation of the consequences of fertility decline, even having circum- scribed the task as I did at the outset.
Although the review that I have undertaken could only be described as cursory and impressionistic, several conclusions have emerged that may stand up to further scrutiny. It is not without interest that several of these conclusions are at odds with the earlier literature on the consequences of fertility decline. While Jones's studies of sectoral effects showed that fertility decline would facilitate attaining educational goals, the quality of education in Latin America, particularly of primary and secondary education, seems to have improved little in recent years, in spite of the decline in the size of birth cohorts. Rather, whatever dividend was generated by lessening demand for school teachers and classrooms seems to have been appropriated by other social sectors, particularly the health sector, where demand was expanding rapidly due both to growth of the adult population, and to increased coverage of social security and public health systems.
If education fared badly in the austere environment of the 1980s, child survival seems to have fared remarkably well. Here, the gains seem to have outstripped whatever improvements might have resulted from changes in the distribution of births by mothers' age and parity, or from improvements in the quantity and quality of medical services. The lesson that I have drawn from this experience is that child survival is more closely tied to fertility than was previously thought, and that the connections are as much social as bio-demographic. Nevertheless, the result is surprising and problematic for those attempting to delineate the social impact of structural adjustment (e.g. Lustig 1991).