Demographic Responses to Community and Family Context

点击翻译按钮获取中文摘要

This project examines how family and community context interact to shape mortality, marriage, and reproduction. In examining the influence of family context, it moves beyond effects of characteristics of parents and other close relatives to focus on the role of kin living in other households and even communities. In examining the influence of community context, it moves beyond consideration of geographic and economic characteristics to focus on community social organization. Inspired by recent developments in the sociological literature on neighborhoods, it assesses whether or not communities where residents were also bound together by kinships ties had more higher levels of collective efficacy and better provision of public goods that resulted in more favorable demographic outcomes. To test theories about the determinants of family change and variation, the research also examines how observed patterns of family and community contextual effects vary in response to secular changes and geographic variation in social, economic, and political context. The research makes use of one of the largest, longest, and most detailed panel data sets ever compiled for an historical and contemporary population - 300,000 individuals who lived in northeast China from 1650 to the present over as many as seventeen generations. The project links existing core data from 1.3million individual-level observations from historical household registers, to records from family genealogies, and retrospective surveys of these same individuals,their foundingancestors who settled the area in the seventeenth century, and their twentieth-century descendants, to produce a single, rectangular file for analysis. These data are supplemented with information at the community, regional, and provincial level on social, economic, and political conditions. Measurements of the influence of family and communitycontext on demographic and social outcomes apply event-history techniques. Fixed- and random-effectmodels are used to account for correlations among individuals at different levels, including theResidential household, descent group, community, and region.