第十八期人口學刊1997.06 出刊
本期目錄
人口再生 ； 生育步調 ； 堆疊與疏散 ； Population renewal ； Fertility tempo ； Piling up and thinning out
中文摘要
近年來由於學者的努力，生育步調對於人口再生的影響已漸累積為人口學的重要文獻；生育步調包括平均生育年齡及其變異兩個成份，兩者雖然在經驗上時常連結互動而變化，理論上卻可各自獨立而影響人口之再生。人口年輪的生育步調改變雖然不致於顯著改變時期生育率的趨勢或走向，對於長期趨勢週圍的時期波動與曲折卻能產生相當大的影響。在我們對於臺灣地區生育率轉型的長期趨勢已有相當瞭解的條件下，對生育步調的解析與研究將能有助於我們進一步瞭解此一長期趨勢之外的時期性變動，在學術上構造一個完整的生育率理論，在應用上則協助我們分析人口趨向靜態度發展時的各項問題。本文使用模擬方法檢討生育步調對出生數量的影響，雖然不見得是比較「精確」的模型，卻較擬似於人口的實際經驗，讓我們可以瞭解並控制模型輸入與輸出數據間的對應關係，深入檢討穩定人口模型所不能直接掌握，只能間接推論的現象。我們特別注意臺灣人口出生數列於光復後所呈現的一前一後兩個峰狀分佈，其峰點距離約為廿一年，顯然關連著人口再生的「記憶與遺忘」過程。1950年代生育峰期出生的人口於1970年代開始生育子女，視其生育率與生育步調而定，以其自有的形式複製其自身出生時所表現的數列形狀，乃有1970年代末期的第二波起伏；由於我們的基礎模擬數列涵蘊著生育步調加快與加緊的事實，第二波起伏的長度乃比第一波要來得短一點，峰度也陡一點，而此一現象乃穩定人口模型的靜態比較所不能直接檢驗的現象。我們的分析也指出第三波起伏於1990年代正在形成中，由於生育率轉型過渡期間的堆疊作用已經消耗殆盡，而晚近臺灣地區的育齡
Abstract
Due to the efforts of Norman Ryder, the influence of fertility tempo on the population renewal process is now-well-known to the demographers. The fertility tempo as formulated is composed of at least two parts: mean age of childbearing and variation in the age of childbearing. Empirically the two parts might be highly correlated in affecting the quantity of birth, theoretically the two parts can have distinct effects on the birth. By manipulating the distribution of age-specific fertilities, this paper simulates and examines the changes in the number of births in Taiwan since 1905. Two humps in birth number after the 2nd World War were indentified in the baseline analysis. The first hump can be attributed to the decline in mortality since 1920 and the decline in fertility since 1951; the growth in population of childbearing age coupled with the decreasing propensity of bearing children have resulted in the growth and decline of birth during the 50`s and 60`s. The second hump was determined a compressed replication of the first one through the renewal process. As a consequence of the continuous fertility compression after the war, the second hump appears to be shorter in length and larger in amplitude than the first hump. Much of the "excess" birth can be eliminated if the variation of age-specific fertilities were held constant to allow for an uninterrupted operation of the "averaging effect".
婚姻狀況 ； 平均餘命 ； 生命表 ； 婚姻選擇 ； 婚姻保護 ； Marital status ； Life expectancy ； Life table ； Marriage selection ； Marriage protection
中文摘要
在過去的文獻中，關於婚姻狀況改變對平均餘命的影響之討論，傾向於使用相對死亡比的方法(Relative Mortality Rate, RMR)來呈現，進而觀察出婚姻選擇作用(marriage selection) 與婚姻保護作用(marriage protection)的效果。本文則嘗試使用另一種方法，即死亡生命表 (life table) 的建構方式，假設出一群經歷不同婚姻狀況變化的人口，觀察他們平均餘命變化的狀況。生命表方法的優點在於，透過假設的設定，可以將其他因素控制住，純粹觀察婚姻狀況變化對平均餘命的影響。研究結果發現，(1)婚姻狀況對平均餘命的影響來看，已婚人口的平均餘命，比其他失婚人口平均餘命要來的高。(2)性別因素對婚姻狀況別平均餘命之影響來看，對女性而言，婚姻所提供的保護作用是持續而穩定，而且要比男性來得好。(3)年齡型態來看，婚姻保護作用對女性而言，一直維持著穩定的影響； 而對於男性，則主要呈現在較年輕的已婚男性人口，50歲以後則效果較年輕者為低，可能與男性受婚姻選擇作用影響較大有關。(4)就婚姻狀況與1981至1991年之間平均餘命之變化來看，女性受到婚姻保護的效果已有減低的趨勢，也顯示未婚女性在獨立生活能力上的增加。相對於過去所使用的RMR方法，使用生命表所得出的研究結果，也可以適當地支持婚姻選擇效果與婚姻保護效果的存在，對於往後關心婚姻狀況與平均餘命變化的研究者而言，提供了一個簡單且值得參考的研究方法。
Abstract
Most of researchers used RMR (Relative Mortality Rate) to study marital status and mortailty trying to reveal the selection and protection effects of marriage on death. This study instead employs life table technique to analyze their effects on life expectancy. Although this study does not intend to differentiate the relative importance between selection and protection effects, through modeling various hypothetical cohort`s marital experiences in life course allow us to control us other factors than the sole effect of marital status. This study finds that:1) remaining constantly in married status results in longer life expectancy than the never-married, ever divorced, and widowed for both sexes; 2) remaining constantly in married status grant more steady protection for females than for males; 3) comparing to males` marital-specific age pattern of life expectancy, it seems that females have more steady gains in life expectancy, which suggests that protection effect of marriage has more intricacy with selection effect among males than females; 4) comparing the change between 1981 and 1991, married females` odds in life expectancy than other marital status decreased. The decrease indicates an increased capability to lead independent life among females.
Migration population growth ； Push-pull theory ； Migration efficiency ； Migration selectivity ； 遷移 ； 推拉理論 ； 人口成長 ； 遷移效率性 ； 遷移的選擇性
Abstract
This paper reviews Japanese and Chinese migration to Taiwan in the pre-war period from the perspective of push-pull theory. Household registration data for 37 years are used to examine the volume, efficiency and selectivity of migration. We have found that migration by Japanese was much greater than that by Chinese, mainly the result of severe restriction on Chinese immigrants. We have also found that migration efficiency for Japanese and Chinese was low. Finally the selectivity of the sex and age compositions of the migrants led to differential population growth for Japanese and Chinese.
中文摘要
本文採推─拉理論的觀點來檢視戰前日人與華人遷徒臺灣的情形。37戶的戶籍統計資料用來檢討遷徒的數量、效率與選擇性。結果發現日人的遷徒量遠大於華人，主因是華人的遷入受到政策性的嚴格控制。另外，也發現日人與華人之遷徙效率都低。再者，性別與年齡組成的選擇性使得日人與華人的遷徒人口呈不同的成長型態。
人力資本 ； 因生育離職 ； 因結婚離職 ； 職業生涯 ； Human capital ； Job career ； Quitted-due-to-birth ； Quitted-due-to-marriage
中文摘要
婦女再就業已經成為一個相當普遍的現象，本研究採用民國八十二年主計處「臺灣地區婦女婚育與就業調查」資料，先描述已婚婦女的就業類型，再以多變項分析來探討「前」「後」職業之「流動」之影響因素。由於婦女工作種類多集中於低階白領和體力工作，離職前與復職後職業的變動難以「流動表」呈現出來，本研究兼採職業地位測量與職業標準分類來呈現「流動」及其「幅度」。在此操作定義下，有超過一半「沒有流動」。經由多變項分析，本研究發現「因結婚離職」與「因生育離職」具有不相似之流動機制；尤其值得一提的是作為「人力資本」的教育程度，雖然都是影響婦女再就業職業改變的重要因素，但在這兩類不同離職原因中卻有不同的效果。簡單地說，在「因結婚離職」的婦女中，教育程度高者傾向往下流動，而在「因生育離職」的婦女中，教育程度高者向上或向下流動都比不流動可能性高。再者，教育的作用對於流動幅度來說也最為凸顯，不但有助於向上流動的幅度，也可減緩向下流動的幅度。總之，教育的意涵隨著離職原因而異，它不在於排除已婚婦女再就業向下流動的可能，而是對於需要向上流動者提供了「助力」，對需要向下流動者提供了「緩降」的功能。
Abstract
Male and female labor forces are different in their patterns of job careers, the latter are generally more likely to be disrupted for marriage and/or birth. That chances of reworking for women (particularly married women) has increased in the past attracts our attention on exploring causes and changes of returning. Our data came from the 1993 wave of the Survey of Women`s Marriage and Fertility in Taiwan. It is shown that around one seventh of married women has experienced returning labor market. Among them, more than half did not move up or down relative to their former occupational status. In multiple covariate analyses, we found schooling is the most essential factor explaining the changes of occupational statuses. Most importantly, the effects of schooling behave differently in the models of quitted-due-to-marriage and quitted-due-to-birth. In short, extra schooling seems to create space of moving down for women of quitted-due-to-marriage, while it functions as a burffer of moving down for women of quitted-due-to-birth.
Causes of death ； Epidemiologic transition ； Cause-of-death structure ； Linear model ； Curvilinear model ； Indirect estimation ； 曲線模型 ； 死亡導因 ； 流行病的轉型 ； 間接推估 ； 線性模型
Abstract
This paper aims to model the relationship between cause of death structure and overall mortality in the Island of Mauritius. The study period extends over an 18 year period from 1969 to 1986 and ten groups of causes of death were studied. During this period, Mauritian mortality for both males and females, underwent a transition from predominnace of infectious diseases to that of degenerative diseases. Three models were fitted to the male and female data, a curvilinear model and a multiple linear model which incorporates the concept of Epidemiologic Transition. It was found that the linear model fitted female data very well while the multiple linear model fitted male data quite well. The curvilinear model generally did not perform well on either the male or female data. From these findings, the paper concludes that in the indirect estimation of cause-of-death structure, information on overall mortality alone will not suffice, supplementary information on the Epidemiologic Transition is needed especially where the mortality pattern is undergoing a rapid change.
中文摘要
本文目的在於建立一個死亡結構導因與整體死亡率關連的模型，研究跨越1969至1986 共十八年，以及考慮十類死亡導因於研究中。 由於在該期間模理西斯的人口死亡導因由傳染疾病為主轉變為非傳染性疾病為主的情形，作者乃建立三種流行病轉型有關之模型 (線性、曲線、多重線性 )，來分析死亡結構。作者發現線性模型之於女性資料以及多重線性模型對於男性資料的適合情形較強；因而建議未來研究，在間接推估死亡結構導因時，光採用整體之死亡率是不足的，應輔以流行病移轉資訊，這種方法對於分析死亡型態快速變遷的地區更為有用。
Out-migration ； Probability model ； Household moment method of estimation ； 外移 ； 機率模型 ； 家戶動差方法之推估
Abstract
This paper presents a probability model for the rural out-migration system at a micro-level. The parameters involved in the model are estimated by the method of moments. The application of the model is discussed and it is fitted to observed data.
中文摘要
本文乃針對微視的農村外移系統提出一機率模型，採用「動差方法」(the method of moments) 推估模型中的參數值，除了以模型檢測資料的適合情形，並討論該模型的應用。