How is Social Time Distributed Over the Age Pyramid? Demographic Baseline Models of Daily Social Interaction in Personal Networks

Christopher S. Marcum, University of California, Irvine

Age differences in the frequency of face-to-face social interactions have long been of interest to life-course sociologists. Most past research in this area has emphasized the role of health status, retirement, and psychological factors in shaping how age influences the amount of time people spend together. Fewer studies have examined propinquity based on demographic baselines. Following Mayhew (1973), I test competing baseline models of age-differences in social interaction in this paper. The results suggest that a substantial portion of social time can be explained simply by household structure, the age distribution, and gender. Data come from the pooled 2003--2007 American Time Use Survey and hypotheses are tested using Butts' (2008) generalization of proportional hazards analysis. Recommendations for further research and implications for social gerontology are discussed