Graduate School of Human Sciences, Osaka University. Nenpou Ningen Kagaku (Annals of Human Sciences). Vol. 22: pp. 17-31. ISSN 0286-5149.
TANAKA Sigeto (tsigeto(AT)nik.sal.tohoku.ac.jp)
[Body of the paper (Japanese PDF; 258KB )]
A secondary analysis of comparable time-use data (arranged by the NHK Broadcasting Culture Research Institute; the survey dates vary from 1985 to 1990) from Japan, Canada, US, UK, the Netherlands, Denmark, and Finland, is conducted by means of a new mathematical procedure to decompose time-use-by-sex arrays distinguishing between effects of marginal distribution and of equalization in lifestyle.
Let a and y denote unpaid housework time by women and by men respectively; x and b denote paid work time by women and by men. Let the marginals be denoted as follows: F=a+x ; M=b+y ; T=F+M ; U=a+y. The index of sex difference in time-use, which is a variant of Yano Masakazu's hubyoudou sisuu [index of inequality], is defined as d=T-2(x+y). Because of the empirical constraint that women's and men's work time (F and M) should balance with each other, d cannot exceed the ceiling of asymmetry C=T-2|F-U|; this C depends on the marginals and presents the structural limit of individuals' option. The appropriate estimator of the prevalence of the gender-equal lifestyle is E=(C-d)/C, a coefficient standardized to be margin-free.
Finally a two-dimensional (E x C) scatterplot of the seven societies is interpreted with the aid of a welfare regime typology by Alan Siaroff (1994):
Duplication in any form without the author's permission is forbidden.
Created at 2000-11-10. Last updated at 2001-08-24.