Page 76 - IJPS-2-2

Transition in risks of higher order births in Nepal: a life table analysis

achieved by age 30 at the latest; almost half of women had the first birth by age 20. Though there
has been a delay in the age at entry into sexual union over time (MOHP, New ERA, and ICF Interna-
tional, 2012; MOHP, New ERA and Macro International, 2007; MOH, New ERA and ORC Macro,
2002), this has had minimal impacts on the age at initiation of childbearing at an aggregate level.
Women marrying at relatively older ages have a tendency to start childbearing earlier after marriage,
mostly for making up. These findings are similar to those obtained in Orissa and Goa, India (Raja-
ram, Rao and Pandey, 1994; Sahoo, 2011). Early and universal first birth followed by universal
second birth is one unique feature of fertility in Nepal.
In accordance with previous studies, this study finds socioeconomic and demographic factors in-
fluence birth interval dynamics. Socioeconomic factors such as women’s education, work status, and
household economic condition, and demographic factors such as age at previous birth, sex of pre-
vious child, and survival status of previous child are significant determinants of higher order births
in Nepal. Although the second birth is almost universal, substantial variation exists in the length of
the second birth interval. A significantly lower risk of the second and higher order births is found
among more educated women, those working in the non-agricultural sector, and those who are from
a better economic background. Similarly, a reduced risk of the second and higher order births is
found among women who start giving birth at older ages and whose first child survived during in-
fancy. This could be the result of declining ideal number of children (MOHP, New ERA, and ICF
International, 2012; MOHP, New ERA, and Macro International, 2007; MOH, New ERA, and ORC
Macro, 2002) and a more effective use of family planning by women who are better educated or
work in non-agricultural sector or are in a better economic status. Declining risk of higher or-
der births over time represents women’s increasing capacity to control fertility. Moreover, women
who are more educated or working in the non-agricultural sector or from sound household econom-
ic condition are more capable of prolonging birth intervals and limiting the number of children.
A substantial decline in fertility in Nepal is largely driven by decline in both the pace and the pre-
valence of the third and higher order births. The rapid decline in fertility in the recent past is attri-
buted to a steep decline in the proportions of women transiting to the third and higher order births. In
the years to come, it is likely that majority of the couples will have two children and they will avoid
higher order births. These results are in the line with those found in Orissa, India (Sahoo, 2011).
Controlling for other factors, there is evidence of some pure temporal effect in fertility decline.
This could be the general experience during the course of demographic transition. A persistent higher
preference for sons is also noticeable: couples want to have a son as early as possible and will either
delay the next birth or stop childbearing after having a son. Similar results are found in Orissa, India
(Sahoo, 2011).
This paper aims to have a better understanding on the fertility transition in Nepal by examining
the dynamics of the age at first birth, the lengths of second and higher order birth intervals, and the
risk of higher order births. This paper also examined the influences of different socioeconomic and
demographic factors on occurrence of the second and higher order births. Life table approach and
Cox proportional hazard models were used to analyze the dynamics of birth interval and risk of birth.
Birth interval dynamics in Nepal were first analyzed when the fertility transition was in its initial
stage with a slower pace (Suwal, 2001). The present paper re-examined the birth interval dynamics
in Nepal for the recent time period, a time in which fertility declined substantially and at a faster
pace, and contributed to fulfill the research gap. Prior studies of Nepal have only used the hazard
model, which cannot show the exact lengths of birth intervals, rather it compares the interval lengths
in terms of relative risks. The novel contribution of this paper is that it has also used the life table
method and depicted the median lengths of birth interval as well as the proportions of women at-
taining next higher order birth after certain period.
When interpreting our findings, the following limitations should be taken into consideration. First,
quantitative findings of this study are based on birth history data, which are subject to content errors
due to recall lapse of the date of event (birth). Second, the analysis is based on cross-sectional survey

70 International Journal of Population Studies | 2016, Volume 2, Issue 2