Gender differences in marriage and mortality for older adults in rural Bangladesh : is widowhood more dangerous for women than men?

by Mizanur Rahman

Purchase

Purchase Print Copy

 FormatList Price
Add to Cart Paperback32 pages Free

This paper uses prospective data from rural Bangladesh to show that economic factors play a dominant role in accounting for gender differentials in widowed versus married mortality ratios for older adults. As in developed countries, males have higher mortality risks in all marital states at all ages. However females have a higher proportionate increase associated with widowhood than do men. These results present a striking contrast to those obtained from developed country data. The disparity in mortality ratios is attributed to the much lower economic status associated with widowhood for females than for males. Access to resources for women is more dependent on marital status and living arrangements than for men. Living arrangements, in particular, the presence of an adult son in the household, has a substantial impact on reducing mortality for older widowed individuals, with the effect being somewhat more statistically conclusive for females than males. Although economic factors dominate, there is some evidence that selection into remarriage after widowhood, on the basis of health status for males, partially compensates for the higher female (widowed/married) mortality ratios.

This report is part of the RAND Corporation Draft series. The unrestricted draft was a product of the RAND Corporation from 1993 to 2003 that represented preliminary or prepublication versions of other more formal RAND products for distribution to appropriate external audiences. The draft could be considered similar to an academic discussion paper. Although unrestricted drafts had been approved for circulation, they were not usually formally edited or peer reviewed.

The RAND Corporation is a nonprofit institution that helps improve policy and decisionmaking through research and analysis. RAND's publications do not necessarily reflect the opinions of its research clients and sponsors.