Determinants of the tempo and quantum of Japanese cohort marital fertility since the 1960s

Abstract

This thesis has clarified the determinants of the period fertility decline in the early 1970s. As a result of decompositions, we found that a large part of the decline in the total fertility rate in the early 1970s was attributable to a drop in the total marital fertility rate. Then 'we investigated the tempo and quantum aspects of Japanese cohort marital fertility since the 1960s which affect the trends in the total marital fertility rate. With respect to the tempo aspects, we examined the trends in and the effects of background variables on wife’s age at first marriage and the first and second birth intervals. As a result, important facts were found regarding the first and second birth intervals. That is to say, the proportional hazards model analysis of the first birth interval reveals that love matches and conjugal families immediately after marriage are accompanied by a longer first birth interval than others even after controlling for other covariates. Marriage cohort also shows a net effect on the relative risk of having a first birth. In accordance with our expectation that, since the 1960s, other things being equal, the spread of parents’ aspiration to children’s high education may have induced a first birth soon after marriage, the marriage cohorts since the mid-1960s demonstrate a shorter first birth interval than the 1961-63 marriage cohort. If we do not control for other background variables, this shortening tendency of the first birth interval is obscured by increases in love matches and conjugal families just after marriage over marriage cohorts which are likely to entail a longer first birth interval. On the other hand, the univariate analysis of the second birth interval clearly shows that it became shorter over marriage cohorts since the 1960s. The proportional hazards model analysis demonstrates that longer first birth intervals, conjugal families immediately after marriage and higher age at first marriage of women are associated with a longer second birth interval. In contrast with the univariate result, once we controlled for other covariates, we found no effect of marriage cohort on the relative risk of having a second birth except in the marriage cohorts of the early 1970s. It was suggested that the change in second-birth timing stemmed from the shortening first birth interval among others and that an attitudinal change towards birth timing in this period was mainly restricted to that of a first birth. A divergent second birth function in the 1970-72 marriage cohort was scrutinized in detail. We found that conjugal families immediately after marriage, wives with low education, wives with white-collar husband, wives with white-collar father and wives with high age at first marriage who married in the period 1970-72 and had a first birth in the period 1972-74 experienced the most pronounced rise in the second birth interval. This might be because their relatively higher sensitivity to a change in socio-economic circumstances than others was released by the highly contrasting experiences around the oil crisis while they married and had a first birth, thereby inducing a delay of second birth. As regards the quantum aspects, we examined the trends in and the effects of background variables on cumulative fertility and found no sizable change in the childcumulation process over marriage cohorts except in the early stage of marriage. Only a small proportion of the variance of completed fertility was explained by background variables. Nonetheless, we discovered a tendency that extended family immediately after marriage was associated with higher cumulative fertility in conformity with the Lorimer- Davis hypothesis. Since a further increase in wife's age at first marriage and another decline in the proportion of extended families immediately after marriage are not likely in the near future of Japan, these two factors will not reduce completed fertility (now around 2.2) any more. Background variables account for only small proportions of the intended and ideal numbers of children and measured intended and ideal family sizes do not show any sign of an increase in the preference for childlessness or the one-child family as observed in some Western countries. Bearing in mind the findings on the tempo and quantum aspects of cohort marital fertility, we decomposed the reduction in the total marital fertility rate in the early 1970s. We found th a t the drop between the periods 1970-72 and 1973-75 wras essentially produced by a change in second-birth timing over marriage cohorts since the mid-1960s. Th a t is, a shift in second-birth timing to earlier second births after marriage since the 1964-66 marriage cohort elicited an increase in the to ta l marital fertility rate in the periods 1967-69 and 1970-72. Then, with a diminution of the effect of old marriage cohorts before the shift in second-birth timing, the total marital fertility rate dropped in the period 1973-75 and this decline was amplified by an ephemeral delay of the second birth in the 1970-72 marriage cohort. We also discovered th a t a reduction in the total marital fertility rate in the period 1979-81 was mainly produced by a delay in the first birth in the 1979-81 marriage cohort. Given th a t the drop in the total fertility rate in the early 1970s is largely a ttrib u tab le to a decrease in the total marital fertility rate, most of the drop in the total fertility rate in this period is also a ttrib u tab le to the change in second-birth timing over marriage cohorts since the mid-1960s and the temporal disturbance in second-birth timing in marriage cohorts of the early 1970s. The shift in second-birth timing since the mid-1960s possibly resulted from reductions of first birth intervals in the same period and the temporal delay of the second birth in the 1970-72 marriage cohort was probably caused by the first oil crisis. It was also suggested th at the delay of the first birth in the 1979-81 marriage cohort stemmed from the effect of the second oil crisis. In relation to the delay of the first birth in the 1979-81 marriage cohort, we explored the possibility of socio-cultural transitions to women's greater independence and self-reliance in present-day Japan. At present there is no drastic and fundamental change in women’s status and a ttitu d e s towards self-fulfilment which is likely to encroach on the firmly established two-child family preference in Japan. Show more