Demographers have been puzzled by whether the stagnation of household formation trends, or the decrease in household headship rate, of young (20-39 years old) married male would continue or not, which appeared from 1975 to 1985 in Japan. The 1990 Population Census result revealed that the headship rates again increased from 1985 to 1990, except for 35-39 years old. Supposing that the change in the coresidence with their parents has played an important role in these complex trends, we have constructed a model which explains the change in headship rate, composed with three variables: the proportion coresiding with parents (c), the headship rate for the coresiding (h_c) and the headship rate for the non-coresiding (h_n), and we have contrived a way to obtain these three variables utilizing the Census tabulations. With this analysis, while the decrease of headship rate during 1975 to 1985 had been brought about by the rise of proportion coresiding and the decline of headship rate for the coresiding, the augmentation from 1985 to 1990 was caused exclusively by the returning decrease in the proportion coresiding. Again, we constructed a model to explain the change in the proportion coresiding (c), composed by two variables : the proportion coresidable (c_a) and the proportion of coresidence realization (c_r), that is c=c_a×c_r. The proportion coresiding is divided into the two factors: the former is purely demographic and the latter is a measure of the propensity to choose coresidence with parents. Using this model, we found that the fall-rise-fall movement of the proportion coresiding from 1970 to 1990 was created by the monotonous increase in the proportion coresidable and the monotonous decline in the proportion of coresidence realization. More specifically, the proportion coresidable had rapidly risen until 1985 deriving from the decrease of sibling size or the fertility decline accompanying the demographic transition more than 20 years before, which caused unexpectedly to many demographers the rise in the proportion coresiding with parents until 1985. However the proportion coresidable reached its upper limit and almost stopped increasing especially for 20's during 1985 to 1990, leading to the reappearing decrease in the proportion coresiding, which was directly caused by the incessant decline in the proportion of coresidece realization or the wane in the propensity to choose the coresidence. This is the cause of the returning increase of household headship rate for the young married male. Understanding these mechanisms, it is natural to anticipate that the increase in the headship rate for the young married male is not temporal and that it will persist until the age pattern of headship rate is almost flat in 2000 which appears similar to those in Western countries.

View full abstract

Based on the data of Survey on Independence of the Elderly, 1991, several analyses were made on independence of the elderly. As it is quite difficult to introduce a strict definition of independence, 3 indices were formulated within the limitation of the data as follows: Index X relating to physical and functional independence ; Index Y relating to mental independence ; Index Z relating to economic independence. Index X was well related to health items as well as dependent on sex and age. Concerning Index Y, there were certain sex differences, but age and household structure did not affect on this index very much. Index Z had strong relationship with the main source of income for elderly households. The independence of the elderly has many aspects and requires researches from multiple approaches. As a multi-dimensional approach to the measurement of independence of the Japanese elderly, the following points were raised, among others, for further consideration : -There were 10 combinations of Indices X, Y, and Z, which had incidence rate of 5 per cent or more among the elderly as a whole. The share of the most common 5 combinations did not change between male and female, nor with age, except male at age group 65-69. -Concerning mental independence of the elderly, there might be some structural pattern which were not related to age or household structure. -Index X was quite well related to all 3 items on Consciousness. Life satisfaction rate was quite high irrespective of Index Y or Index Z, but age consciousness was found to be strongly related to Index Y and economic satisfaction with Index Z. -As for the characteristics of female elderly living in one-person household, such combinations of indices as (b, a, a), (c, b, b) appeared more often and (b, b, c) appeared less compared to the other female elderly (see 3.2 for notation). -An artificial age was calculated for each individual elderly from his or her actual age and activity level. The rate of artificial age falling in the same age group as actual age decreased rapidly with age increase from 55 per cent at age group 65-69 to 38 per cent at age group 80-84.

View full abstract