Determinants of living arrangements and changes in living arrangements of elderly Singaporeans: A cross-sectional and longitudinal analysis

As a result of the rapid demographic transition that Singapore underwent in the context of socioeconomic development, the city-state currently has one of the fastest aging populations. In Asia, where the main source of elderly support continues to come from the family, coresidence with adult children is regarded as a prime indicator of elderly well-being.;The research is guided theoretically by: (1) the rational choice model which takes into account opportunities, resources and needs, and culture and preferences; (2) the person-environment model (Lawton, 1981) which takes into account lifecourse events through the aging process. The data come from two panel waves over a four-year period: The 1995 National Survey of Senior Citizens in Singapore and Transitions in the Health, Wealth and Welfare of Elderly Singaporeans, 1995-1999. The sample is stratified by parents' marital status, with the characteristics of the spouse included in the married parents' analyses. I also distinguish between living with married and unmarried children since the direction of support often differs by the children's marital status. Multinomial logistic regression methods using a stepwise modeling approach are used to analyze the data.;The results indicate that parents' marital status is an important determinant of their living arrangements. The effects of parents' demographic characteristics are strongly mediated by the marital status and gender of the children. Parents' homeownership significantly reduces the likelihood of living with married children and increases the likelihood of living with a spouse only. As a proxy measure of their resources, parents' work status does not have an effect on both married and unmarried parents' living arrangements. Health-related conditions exert significant effects on the living arrangements of unmarried parents only. In comparison to other factors, culture and preferences, measured by race and education, are weakly associated with the living arrangements of Singaporean elderly. The longitudinal analyses which are largely similar to the cross-sectional analyses provide additional evidence in support of the causal effects of these variables on living arrangements. As for changes over time, baseline living arrangements strongly predict follow-up living arrangements of older Singaporeans.