近年來台灣地區育齡婦女總生育率由民國64年的2.8人，降為民國73年的2.1人。促使婦女總生育率下降的因素是什麼？Kinsley Davis :& Judith Blake於五十年代最先提出生育模式（Reproductive Model）的理論：婦女生育的下降係直接受生物與行為變項的影響，而社會、經濟、文化及環境等因素，則透過生物與行為因素，間接地影響婦女的生育。惟為了探討這些關係，得各項變項加以量化，常因此產生了一個高度複雜化的生育模式，實際研究上是相當地困難。故而一般學者與專家均著重於探討社會、經濟學因素對生育的影響。於七十年代末期，John Bongaarts針對Kinsley Davis 及 Judith Blake（1956 ) 的生育模式，提出了一個相當完整與簡單的模式，以分析直接影響因素和生育高低間的關係。本文即嘗試運用Bongaarts所提的模式，來探討影響台灣地區婦女生育率下降的直接影響因素，何者影響最大？影響程度如何？又在政策的推行與制訂上具有何種特列的涵義及有何可參考之處？況且我國於民國74年公布實施優生保健法放寬施行人工流產條件，則該法的公布，對生育下降的影響文如何？亦可利用此次研究，瞭解法公布前婦女施行人工流產對生育力的影響，以建立起探討該法公布後影響力變化的依據。 本主運用行政院衛生署委託台灣省家庭計畫研究所辦理的「72年及73年台灣地區有偶婦女人工流產實施狀況調查計畫」所搜集的資料，以Bongaarts的理論，加以分析，評估四個最主要的直接影響因素--結婚比率、產後哺乳不孕期、使用避孕方法及施行人工流產，對婦女總生育率下降的影響情形。結果顯示，以Bongaarts模式估算出台灣地區總生育率的估計值為1.96，較觀察值 (即樣本的總生育率) 2.03 ，低了0.07，其差異仍在Bongaarts的實證研究差異 (0.36 ) 範圍內，但設四個指數對組生育率的影響為一，對各指數取對數 ( log ) 值，可明顯地看出台灣地區四個指數對總生育率的影響情形。顯示避孕指數對總生育率的下降影響最大，為47%。其次為結婚指數，其影響比倒為41%。至於另兩個指數，人工流產指數與直接哺乳不孕指數，對總生育率下降的影響相當小，前者為11%，後者僅為1%。 以不同行政區域來看，其估計的總生育率：七大都市為2，01，縣轄市與鎮為1.96，鄉村為2.07，與其觀察值 (1.77，2.02及2.31 )比較，略者差異。不同地區別的四個指數的影響，均仍以避孕指數的影響最大，七大都市為45%，縣轄市及鎮為46%，鄉村為50%；其次為結婚指數，其影響比例：七大都市為44%，縣轄市與鎮及鄉村均為40%。人工流產指數以縣轄市與鎮的影響較大為13%，鄉村地區影響最小為9%。產後哺乳不孕指數的影響最為微小，對七大都市的總生育率完全無影響，另兩個區域的影響力也僅為1%。

In Taiwan the total fertility rate decreased from 2.83 per woman in 1975 to 2.05 in 1984. The decline appeared accelerated during the recent years.. Promotion of family planning program has been implemented since 1964. IUD, oral pill, condom and sterlization are the methods provided by the program. Induced abortion has only been legalized based on specified grounds in July 1984. To establish the base line data and to evaluate the impacts of induced abortion on fertility before and after promu1gation of the Law, an induced abortion prevalence survey was conducted by the Taiwan Provincial Institute of Family Planning in 1985 and sponsored by the Department of Hea1th, the Executive Yuan. 12,516 samples of married women age 15-49 were covered in the survey. The sizable sample is significant to represent Taiwan area. Bongaarts’ model of proximate determinants of ferti1ity is applied on the survey data to analyze the relationships among total fertility rate, total fecundity, and the indices of the proximate determinants on non-marriage, contraception, induced abortion, and lactation. Using the Bongaarts’ model, the estimated total fertility rate is 1.96 which is slightly lower than that of 2.03 of the observed value and the total fecundity of 15.3 which is estimated by Bongaarts’ & Potter’s. The formulated proximate determinants of total fertility rate, total fecundity, and indices of proximate determinants of non­marriage, contraception, induced abortion, and lactation are shown as follows: TFR (total fertility rate) = TF (total fecundity) × Cm ((index of non-marriage) × Cc (index of contraception) × Ca (index of induced abortion) × Ci (index of lactational infecundability) 1.96 = 15.3 × 0.43 × 0.38 × 0.80 × 0.98 By the logarithm method to portion the difference between total fecundity and total ferti1ity rate among the four proximate determinants, the index of contraception (0.38) appears to have the greatest (47%) effect on the decline of TFR; followed by proportion of married which counted for 41 %. As to the other two indices, induced abortion and lactational infecundability had slight effect, of 11 % and 1 % on the dec1ine of TFR, respectively. As regards to the different urbanization areas, the estimated value of TFR of large cities is 2.0l per woman,1.96 for urban township, and 2.07 for rural areas. Compared with the observed value of 1.77,2.02, and 2.31, the differences are small. The effect of four indices among the different urbanization areas, contraception has also the greatest effect on fertility dec1ine. The effect of 45% for large cìties, 46% for urbantownship, 50% for rural areas. The reason for higher effect in the rural areas than the large cities, is due to the fact that the woman in these areas use more effective method than the woman in other areas. As regards to the effects of the proportion of married on the fertility of three different urbanization areas, they are 45%, 40%, and 40%, respectively. As of induced abortion, the effect is 13% inurban township and 11 % in large cities. This fact may be due to underreporting in large cities. Other studies indicated that residents in large cities have more induced abortions. Lactation infecundabi1ity had no effect on decline of fertility in large cities but it had a very small impact of 1 % in urban township and rural areas.