年輪生育率分佈改變對時期生育率的影響

The Effect of Cohort Fertility Schedule Changes on Period Fertility

王德睦(Te-Mu Wang) ； 董宜禎(Yi-Jhen Dong)

《人口學刊》 48期 (2014/07) Pp. 1-41

https://doi.org/10.6191/JPS.2014.48.01

摘要

臺灣的時期總生育率（total fertility rate）於2011年已降至1.06，僅及人口替換水準的一半。事實上，除了實質生育率下降外，延後生育也是造成生育率如此低的重要因素，文獻稱為步調效果（tempo effect）。然而婦女生育年齡有限，生育不可能無限延宕，當生育不再延後，時期生育率就會回升。為了去除生育延後對時期生育率的影響，Bongaarts and Feeney（1998）提出校正方法，他們僅考慮平均生育年齡的改變，忽略生育年齡變異數的效果；Kohler and Philipov（2001）納入變異數的變動以補足其缺失，目的均止於步調去除後的生育水準。本文乃進一步使用臺灣婦女的生育資料，由模擬方式分別估計年輪平均生育年齡改變與年輪生育年齡分配型態改變各自對時期總生育率的影響。首先設定各年輪完成生育率均維持觀察到的量，生育年齡分佈則維持「比較基準年輪」之型態，比較歷年時期總生育率的模擬值與觀察值之差異，得出年輪生育率分佈改變的總效果。再設定年輪的平均生育年齡為觀察的平均年齡，分配型態同於「比較基準年輪」，其與第一個模擬之時期總生育率相較後，即為一般所稱的步調效果；其與歷年觀察的總生育率之差異，則為分配型態改變的效果。藉此釐清年輪的平均生育年齡改變與生育年齡分配型態改變各自對時期總生育率的效果，用以補足過去文獻的缺失，乃具有人口學應用及學術參考價值。

關鍵字

完成生育率 ； 時期總生育率 ； 步調效果 ； 時期性效果 ； 時間效果

並列摘要

The total fertility rate in Taiwan has declined since 1951, reaching half of the replacement level in 2011. The secular decline is obviously real, but a ＂tempo effect＂ has also worked to bring it down. Since fertility cannot shift indefinitely to later ages of childbearing, the total fertility rate can be expected to rise once the shift has become saturated. Bongaarts and Feeney (1998) proposed a model to correct this terminating effect of shifting age pattern on total fertility, but it appears to address only the shifting mean age of childbearing. The model neglected the effect of shifting variance of the age distribution on total fertility rate. To redress the problem, Kohler and Philipov (2001) developed a way to assess the effect of removing variance on total fertility rate. The remedy appears to focus on the outcome of the tempo effect, but pays little attention to the shifts in the age pattern of fertility itself. This study employs a simulation approach to capture both the effects of shifting mean age and variance of the age distribution on total fertility rate simultaneously. We first compute cohort completed fertility rate based on the observed age-specific fertility rates. We then adjust the age pattern of cohort fertility assuming a certain age pattern, and simulate the total fertility rates accordingly. The differences between the simulated total fertility rates and the actual total fertility rates over time are then considered as the results of shifting age pattern. We further adjust the cohort fertility to the averaged mean ages of childbearing while keeping the age pattern constant; the comparison between this new simulation and the earlier one amounts to the well-documented tempo effect.

並列關鍵字

completed fertility rate ； total fertility rate ； tempo effect ； period effect ； timing effect

參考文獻

劉一龍、李大正、王德睦(2008)。調整生育步調對臺灣總生育率的影響。臺灣社會福利學刊。6(2)，25-60。

楊靜利、李大正(2008)。臺灣人口資料之編製與調整：1905-1943與1951-1997。調查研究―方法與應用。23，119-154。

王德睦、劉一龍(2008)。臺灣總生育率再分析―ACF法的運用。人口學刊。36，37-65。

Hajnal, J. 1947. “The Analysis of Birth Statistics in the Light of the Recent International Recovery of the Birth-Rate.” Population Studies 1(2):137-164.

Barbi, E.(ed.),Bongaarts, J.(ed.),Vaupel, J. W.(ed.)(2008).How Long Do We Live? Demographic Models and Reflections on Tempo Effects.Berlin, DE:Springer.