由於台灣近年來生育率有逐年下滑的趨勢，形成少子化的現象，少子化代表著未來人口可能逐漸變少，對於社會結構、經濟發展等各方面都產生重大的影響，因此此一課題值得深究。 本研究以文獻分析及深度訪談法來進行，首先利用文獻資料來分析影響生育意願的相關原因，包括人口結構變遷對總體經濟的影響、生育行為的相關理論、現實經濟的因素、婦女勞動參與等，並歸納整理後作為深度訪談的基礎。深度訪問以大台北地區12位已婚婦女為對象，藉由深入訪談來了解箇中原因，歸結出台灣目前超低生育率的原因所在。 本研究結果發現，目前婦女生育意願低落的原因，主要是第一：傳統養兒防老的觀念式微，女性生育自主意識的崛起；第二：受教育延長、晚婚是催化高齡少子的趨勢之一；第三：現實經濟的考量下，加上養兒不易，導致多數家庭因此將生小孩的數量轉往養育的品質；第四：家庭分工的不平均也是降低婦女生育的意願的重要原因之一；最後現階段的托育政策與制度的不完善，加上生育補助有限，誘因不足之下，因而無法提高婦女的生育意願。 依據研究發現本研究建議(一)強化家庭功能，以增加婦女的社會支援；(二)提高育兒補助並視嬰幼兒為未來國家的資產，以減緩少子化的趨勢；(三)政府應以人口優質及避免人口短缺的方向努力；(四)政府應積極擬訂鼓勵男性參與育兒及家務政策，並將養育責任，由兩性來共同承擔，以達到美滿家庭的目標；(五)加強改善婦女的經濟地位，鼓勵並積極輔導婦女就業及強化專業職能的訓練，不致因婚育而影響其勞動參與，以提高其生育意願；(六)政府應正視目前超低生育率的現象，設立專責機構並參考國外(如日本或法國)生育政策及托育制度，以提高國內生育率。 關鍵字：少子化、超低生育率、生育行為、生育意願 / In Taiwan, the rate of fertility has been continually in decline in recently year, which is called ‘the phenomenon of declining birth rate.’ It signifies a less future population and impacts on every aspects of our social structure and economic development. This, it is worthy of probing. This research is conducted with the method of literature review and in-depth interviews.It first analyzes via literature review those factors influencing the will of bearing children, such as the impact of the structural change of population on the macro economy, the theory of fertility behavior, considerations of livelihood, and the women participation in labor, etc. All those factors are inducted and sorted as the basis of in-depth interview. Then, the in-depth interview is conducted on 12 married women, penetrating on the cause of extremely low fertility rate in Taiwan. This research finds the reasons to cause the low rate of fertility are as below: First, the traditional idea is decline currently to have children caring the agedness. Meanwhile, the females are learning the independent consciousness to decide how many children they want in life. The second: high grade education & late marriage is one of the trends of declining birth rate with elder females. The third: taking consideration into the family’s economy as well as hard teaching on child, most families turn to focus on life quality instead of having kid's amount. The fourth: the heavy loading of housework on females is also the reason to reduce the will of bearing children. Finally, not only the current child care policy/system but also the limited subsidization by the government is not perfect to meet the family’s requirements; which are unable to promote the will of bearing children. Based on such research findings, I come up with six policy recommendations as follows to encourage our country’s bearing rate：(一) to enhance family function according to the research, with increment women of social support；(二) look after the kids as the country’s property to promote the subsidization of bearing children to reduce the declining birth rate；(三) government should forward to superior quality population, and to avoid the population runs short；(四) the government should actively draw up some bearing plans to encourage male’s participation (sharing evenly) on child care, housework & teaching responsibility in family; which is aimed to build a sweet home for couples；(五) to improve the female’s economic position. And, to encourage/guide women to work with complete working training for proper job arrangement, which will encourage their will of bearing children when woman gets independent incomes；(六) the government should notice the serious declining birth rate currently – i.e. not only to establish some responsible organizations, but also to refer successful growing policy and child care system from other countries (such as Japan, France) in order to raise the fertility rate in Taiwan..

人口統計學家在傳統上習慣將人口的種種變化視為時間的函數，皆試圖以決定型（deterministic）的函數來刻劃，例如：1825年Gompertz提出的死力法則、1838年Verhulst以羅吉斯函數描述人口成長。近年則傾向於逐項（item-by-item）分析各種可能因素，例如：1992年Lee-Carter提出的死亡率模型、目前英國實務上使用的Renshaw與Haberman(2003)提出改善Lee-Carter模型的Reduction Factor模型、加入世代（Cohort）因素的Age-Period-Cohort模型等。但台灣地區近年來生育率與死亡率皆不斷下降，且有隨著時間而變化加劇的傾向，使得以往使用的模型不易捕捉變化。 本文以另一個角度思考生育與死亡變化，將台灣人口視為一隨時間變化的動態系統，使用微分方程來刻劃，找出此動態系統的背後所隱含的規則。人口動態系統的變化，主要來源是出生、死亡與遷移，在建模的過程中，我們先各別針對其中一項，在其他條件不變的情況下，以常微分方程建模，之後再同時考慮各項變動，以偏微分方程建模，找出台灣人口變化的模型。在本文中，我們先介紹使用微分方程模型分別配適與估計出生與死亡。 由台灣地區人口統計資料顯示，不論總生育率或各年齡組的死亡率都有逐漸下降的趨勢，但是每年之間的震盪很大，因此我們提出「二次逼近法」，從出生或死亡對時間的變化率與曲度來估計生育率與死亡率，對於此種震盪幅度較大的資料，可以得到頗精確的估計。唯在連續幾年資料呈現近似線性上升或下降處，非線性的模型容易出現較大的估計誤差，針對此問題我們也提出一些可能的修正方法，以降低整體的模型誤差率。 / Conventionally the change of population is considered as a function of time and described by using deterministic functions. The well-known examples are Gompertz law of mortality (1825) and Verhulst’s logistic growth model (1838). Recently demographers favor stochastic models when analyzing factors in an item-by-item fashion. Since 1992, Lee-Carter model is a most commonly used stochastic model in demographic studies. But empirical studies indicate that the rapid declines in both fertility and mortality rates are against the assumptions of Lee-Carter model. In this study we treat Taiwan population as a dynamic system which changes over time and characterize it by differential equations. Since the changes are from birth, death and migration, we first separately build models using ordinary differential equations. Afterwards the model of Taiwan population can be built by using partial differential equations considering the three main factors simultaneously. Total fertility and age-specific mortality rates in Taiwan decline over time but with shakes between years. Consequently we propose‘parabola approximation method’and apply it to velocity and acceleration of birth or death to solve the differential equations of Taiwan fertility and mortality. Empirical study shows the method allows us to get accurate estimates of mortality and fertility when the data change a lot in a short period of time. But we found the model may over-fit the data at some time point where the function does not seem to be very continuous.

本論文在實質與貨幣內生成長模型中探討政府租稅的效果，而其中又以稅制改革政策為主角。 在第二章中，我們將內生生育率的概念引入了Romer (1986)的實質經濟成長模型，並說明當政府在維持稅收中立的原則下將所得稅制轉換為消費稅制時，將可能對經濟成長與社會福利有所傷害。而後我們也提供了一些數值模擬以支持我們的論點。至於第三章，我們將勞動供給內生的概念引入一個有預付現金限制(Cash-in-Advance)的貨幣經濟成模型，在本章中我們得到了所謂「Mundell-Tobin Effect」與「消費稅中立性」的成立與否，將取決於政府稅收的用途；接著我們把焦點放在資本的生產外部性與消費稅所造成的扭曲上並導出最適貨幣政策；最後，在維持政府的支出水準下，我們得到將消費稅制轉換為通膨稅制時，將對經濟成長有正面的效果。 / The dissertation provides a theoretical framework to investigate the effects of tax policies, especially the tax reform, in real and monetary models of endogenous growth. In Chapter 2, by shedding light on the endogenous fertility choice, we set up a simple Romer (1986)-type endogenous growth model and show that, in a departure from the existing literature, a switch from a decrease in income tax rate to an increase in consumption tax rate so as to ensure a revenue-neutrality could be harmful, rather than favorable, to both growth and welfare. In addition, we also conduct a simple numerical analysis to investigate the conditions in which the negative effect on growth and welfare occurs. As to the monetary model, an endogenous growth model with endogenous labor-leisure choice and cash-in-advance (CIA) constraint which is only imposed on consumption is established in Chapter 3. Through the model, we found that the Mundell-Tobin effect and the validity of consumption tax neutrality depend on the usages of tax revenue. Next, focusing on the distortions due to the production externality of capital and consumption tax, the optimal monetary policy is also derived. Finally, we show that a switch from consumption taxation to inflation taxation to finance a given stream of government expenditure, namely tax switch, enhances economic growth through the increase in labor supply in a CIA economy and the qualitative equivalence between MIU and CIA approaches is still valid.

台灣地區在2002年時總生育率已降至1.33，而淨繁殖率為0.63，此種低於替代水準的生育率如果繼續下去，台灣地區的人口成長將於20-30年後，達成人口零成長並轉為負成長；生育率持續降低將造成人口老化，對國家經濟造成負面影響。 而影響生育率的因素很多，鑑於現代家庭在子女養育費用上的支出日益增加，尤以教育費用佔養育成本之比例最大，中低收入的家庭籌措教育費用的負擔，可能比生活開支更為沉重；尤其近年來政府廣開大學之門，但高等教育的經費並沒有比例的提升，因此不論公立或私立學校，從政府所得的補助相形減少；而大學學歷已成為基本的學位，可以預期未來幾乎每個人都會念大學，因此當大學學雜費調漲，將使父母養育子女成本增加，所以要思考是否因為養育成本的提高，而降低了生育子女的意願，使得台灣地區生育率年年下降。 本文的研究目的，主要在探討佔教育費用相當大比重的大學學雜費調漲，對台灣地區生育率的影響，並就其他影響台灣地區生育率的主要因素加以分析。以下為本論文之主要內容：第一章緒論，介紹本文研究背景與動機、研究目的與研究架構及研究限制。第二章文獻回顧，探討影響生育率因素的相關文獻，並就大學學雜費之文獻加以分別整理並予以說明。第三章，分析台灣地區生育率與大學學雜費調漲之變化；為了研究大學學雜費調漲對台灣地區生育率之影響，必須對近年來台灣地區生育率之變化及大學學雜費制度之變化作進一步之了解。第四章是研究方法，在本章中，將介紹本論文的資料來源、研究方法與實證模型及變數說明，以檢視大學學雜費調漲對台灣地區生育率的影響。第五章，就實證結果加以分析說明，第六章是本研究之結論與政策建議。 本文以1991年至2002年台灣地區23個縣市之追蹤資料為研究對象，經由固定效果模型的估計結果發現，大學學雜費對台灣地區生育率的影響有顯著的負相關；而其他影響台灣地區生育率的主要因素，如台灣地區各縣市家庭所得、失業率（尤以男性失業率為顯著）、婦女教育程度等均呈現顯著的負向影響，台灣地區嬰兒死亡率亦呈現負向影響但並不顯著，而婦女25 – 39歲年齡組群佔該縣市15歲以上婦女總人數比例，對台灣地區生育率的影響，呈現顯著的正向關係。 依本研究結果顯示，大學學學雜費是影響台灣地區生育率相當重要的因素，且教育學費與提升生育率皆屬於政府的政策，政府應該高度正視此問題並有所作為。至於控制其他因素後之地區特質效果，`顯示台灣地區之生育率，以東部地區為最高，南部地區最低；各縣市之特質效果則以新竹縣為最高、高雄市最低，可能與原住民比例、人口外移、城鄉差距或其他人文因素有關。因此，如何縮小城鄉差距,平衡人口分佈,使經濟均衡發展,也是值得政府思考與正視的問題。 關鍵詞：大學學雜費、生育率、固定效果模型、台灣地區

目前超低生育率現象，已對整體社會的走向、國家經濟的發展、生活思維的方式等都產生了相當大的影響。換言之，生育不僅是婦女本身的問題，它與國際社會的動態、經濟的全球化、企業的國際化、現代的子女教育、社會的性別教育、老年人的護理等都有著相當緊密的關係。因此，本研究的主要目的，在於探討臺灣各縣市房地產價格對生育率的影響，並瞭解臺灣生育率持續下降的癥結所在，找出可能影響生育率下降的主要因素加以分析。 本文利用臺灣23個縣市別的追蹤資料，涵蓋期間為2000年至2008年，採雙因子固定效果模型進行實證分析。估計結果發現，在控制其他變數不變之下，在生育率遞延二期時，臺灣各縣市平均地價對生育率為負影響，且在生育率遞延三期時更為顯著，兩者之間的資源排擠效果更為強大。而其他影響臺灣生育率的主要因素，如：各縣市農業人口比例、各縣市粗結婚率、與外國人結婚人數比例皆呈顯著正影響；臺灣各縣市平均地價、原住民人數比例及婦女勞動參與率皆呈顯著負影響。 依本研究結果顯示，房地產價格是影響臺灣生育率相當重要的因素，政府應需高度正視此問題並有所作為。建議未來政府在訂定國家發展政策時，除需研擬更多提升生育率的策略，如：降低父母生育子女的機會成本、改善學前幼托體系，以及加強減輕家庭育兒壓力的宣導工作等，並可增加控制「臺灣房地產價格」及避免房價炒作之策略，以促進國家永續發展及增強國際競爭力。 / The current low fertility rate has been a great impact on our entire social trend, the development of national economy, and the way we think. That is, fertility is not only women’s consideration, but also an issue connected closely to international society, economical globalization, internationalization of business, children education, sex education, and nursing care of the elders. As a result, the main purpose of this paper lies on the effect of average land price on fertility rate in Taiwan. This article uses the panel data of the 23 counties and cities of Taiwan from 2000 to 2008, and does the evidence-based analysis by two-factor fixed-effects model. We found that the average land price has conspicuously positive effect on the fertility rate in Taiwan. While other main factors affect the fertility rate in Taiwan such as the proportion of farm-workers, crude marriage rate, and the proportion of marriage with the foreigners have conspicuously positive effects on it; on the other hand, female labor force participation and the proportion of aborigine have conspicuously negative effects on it. According to our study, average land price is the most important factor which affects the fertility rate in Taiwan, and the government should face the problem and do something to it. When making policy, it is suggested the government may take into consideration of how to control the housing price in Taiwan and avoid housing price speculation.

本文的主要研究目的，在於將文化因素（即龍年與虎年效果）納入生育行為的考量，以重新審視台灣生育率與婦女教育程度之間的Granger因果關係。利用台灣1952年至1994年資料進行實證的Cheng and Nwachukwu（1997）及Cheng（1999），其結論指出台灣教育程度與生育率之Granger因果關係並不顯著。本文將以該文章之實證模型為基礎，並加入代表龍年效應與虎年效應的虛擬變數以建立本文模型。 首先根據Cheng（1999）的資料與變數，利用台灣1952年至2005年之年資料，建立一個包含粗出生率、教育程度大專以上的比例、女性勞動參與率、與實質經濟成長率等四個內生變數，以及代表龍虎年效應的兩個虛擬變數之VAR模型。並以Toda and Yamamoto（1995）提出的Granger因果關係檢定，檢定台灣教育程度與生育率之間的因果關係。隨後，為增進估計的有效性，本文利用台灣地區1978年至2005年的季資料，共112個樣本進行實證研究。變數定義方面，將粗出生率替換為一般生育率，教育程度大專以上的比例替換為育齡婦女大專以上的比例。並根據上述建立之VAR模型，進行Granger因果關係檢定。 經由本文的實證研究發現，將龍年效果與虎年效果納入考量後，台灣婦女教育程度的提昇會Granger影響生育率的下降。亦即台灣婦女教育程度日益提升，是解釋生育率下降的重要因素，此結果與Cheng（1999）所提出的結果並不一致。此外，龍年與虎年對台灣的生育率分別有顯著正向與負向的影響。 / This paper tries to revisit the null hypothesis of Granger no-causality between female education and fertility rate in Taiwan, by considering the culture factors captured by Dragon and Tiger Years which might influences the fertility behavior. In addition, this study compares the primary finding with the result proposed by Cheng and Nwachukwu (1997) and Cheng (1999) which are that female education does not affect fertility rate in Taiwan. According to Cheng (1999), official time series yearly data from 1952 to 2005 provided by Taiwan government are used first, and quarterly data from 1978 to 2005 are required to improve the efficiency. This study models a 4-Variables VAR and applies Granger no-causality test proposed by Toda and Yamamoto (1995). The primary finding of this study is that there is a negative causality from female education to fertility rate in Taiwan, which is inconsistent with conclusions in Cheng (1999). In addition, culture factors do play a very important role in fertility behavior in Taiwan.

生育率的降低是影響台灣地區近年來人口老化的顯著因素，因其變化幅度通常高於死亡率，對人口結構的影響較大。過去台灣地區生育率研究多為整體生育(如：總生育率、年齡別生育率)趨勢的模型，較少探討台灣各地區的特色。為能更深入瞭解台灣生育行為變化的特性，本文將生育率的研究層面由整體的資料，延伸至全台灣地區的各鄉鎮市區（不含離島地區有350個鄉鎮市區），希冀能更精確地找出與台灣地區生育率持續下降的相關因素。本文分為兩個部份，以鄉鎮市區的年齡別婦女生育率與年齡別有偶婦女生育率為研究對象，資料時間為1991、1992、2001、2002年：第一部份探討各鄉鎮市區的生育率數值間是否存在空間相關性，並進一步瞭解生育率較高（或較低）的地區是否有聚集的現象。第二部份則套用空間迴歸模型探討與生育率數值有關的因素（例如：人口密度、教育程度等），更精確且客觀地提供生育率未來趨勢的建議。 關鍵字：生育率、人口老化、空間統計、空間群聚、空間迴歸 / Both the fertility rates and mortality rates, especially the fertility rates, have been experiencing dramatic decreases in recent years, and the population aging thus has become one of the major concerns in Taiwan area. In order to identify the factors that are related to the decrease of fertility rates, unlike the previous works that deal with the aggregate national data, we will study the fertility pattern in township level. We will use the data of age-specific fertility rates and total fertility rates in 1991, 1992, 2001, and 2002 in 350 townships of Taiwan area. This study will be separated into two parts. First, we shall explore if there is spatial correlation among 350 townships of Taiwan area and detect if there are spatial clusters for higher fertility townships. The second part of this project will be focused on the spatial regression model. We will use this model to determine the factors that are highly correlated to the dropping of fertility rates. Key Words: Fertility Rates, Aging Population, Spatial Statistics, Spatial Clustering, Spatial Regression

由於台灣地區的生育率變化較大，之前的研究發現其他各國的生育率模型不見得適用，亟需建立可反映我國國情的生育率模型。本文引用台灣、日本、荷蘭、美國（亞洲、歐洲、美洲）等經濟發達國家的出生資料，配適包括Gamma、Lee-Carter、主成份分析、年齡組個別估計法、擴散模型等較為常用的模型，比較這些國家配適結果的異同。分析發現，如果要預測總生育率，台灣、日本、美國都是以年齡組生育率個別配適擴散模型的結果最佳，荷蘭則是年齡組個別估計法；在年齡組生育率的預測方面，台灣、日本、荷蘭都是以年齡組生育率個別配適擴散模型最好，美國則是以年齡組個別估計最好。此外再從相對穩定性或相對效率的角度來評判，一樣是以年齡組生育率個別配適擴散模型或年齡組個別估計的總生育率預測結果最佳。最後還觀察到台灣地區和各縣市的有偶婦女比例和生育率呈正向關係，平均生育年齡和生育率呈反向關係，各縣市在有偶婦女比例、生育率、平均生育年齡的變化並不一致，各年齡組有偶婦女比例和生育率的改變也不盡相同。

近年來國人晚婚、不婚、晚育或不育的現象持續惡化，衍伸出高齡少子化所帶來之人口結構惡化已是國家安全層次議題。本研究以實證方式探討影響台灣地區生育率之主要因素，並著重住宅價格（房價所得比）、住宅交易數量（建物買賣移轉棟數）及存量（自有住宅率）之變化與總生育率之關聯性。探究近年我國房價變化、持有自有住宅比率及住宅市場之交易活絡程度是否會影響家戶單位生育決策。 研究針對西元2002年至西元2015年間台灣地區20縣市共280筆追蹤資料(Panel data)，在控制時間效果下運用傳統最小平方法、固定效果模型及隨機效果模型對資料進行迴歸分析。研究結果顯示：在考慮生育率落後一期情境下，同時控制時間及區域效果模型中發現：房價所得比與總生育率呈現顯著正相關，建物買賣移轉棟數與總生育率呈現顯著負相關，自有住宅率雖與總生育率為正相關但未達顯著水準。而其他影響總生育率之主要變數：女性受高等教育比率及粗結婚率與總生育率成顯著正相關。失業率與總生育率則為顯著負相關。研究結果可發現，可能由於我國高度自有住宅率之特性，房價上漲帶來之財富效果是導致房價所得比與生育率呈現正向關係之成因。建物買賣移轉棟數與總生育率呈現負向關係，研判來自於家戶購屋初期通常需付出一筆為數不小之頭期款，短期間內將耗用家戶較多經濟資源，進而排擠家戶單位短期生育決策。 / In recent years, more and more Taiwanese tend to marry at a later age, or remain unmarried, or bear children at a later age or not at all. The problems of population aging and low birth rate have led to the worsened population structure, which has become an issue of national security. This empirical study investigated the factors that contributed to the present low birth rate in Taiwan. The study emphasized the relationships between housing price to income ratio, the number of housing transferred, homeownership rate and total fertility rate (TFR). This study also tried to investigate how the aforementioned three factors affect the childbearing decision of family units. The study focused on a panel data set with 280 samples that was collected across 20 cities and counties in Taiwan from 2002 to 2015. Ordinary least squares, fixed effects model and random effects model were used, and time effect are controlled in all models. The results showed that, after controlling for the time and regional effects, the housing price to income ratio had a positive relationship with the lag TFR. The number of building transferred had a negative relationship with the lag TFR. The homeownership rate had an insignificant negative relationship with the lag TFR. Other factors also affected the lag TFR. For example, the ratio of females who received higher education and crude marriage rates had a significant positive effect. The unemployment rate had a significant negative relationship with the lag TFR. The results might imply that, given the high homeownership rate in Taiwan, the rise in housing price increase “wealth effect’’ which is the reason for the positive relationship between housing price to income ratio and lag TFR. The reason for the negative relationship between the number of building transferred and lag TFR could be the huge financial burden of down payment when a family purchases a house. This burden takes out a huge portion from the budget, and therefore family units might decide to delay having children for a short period.

人口政策是政府的重要政策之一，而總人口數則是政府制定政治、經濟、社會及文化發展計畫之主要參考依據，因此如何準確地預測未來的總人口數就成為政府相關部門重要的課題。 本論文試圖為台灣地區總人口數建立時間數列預測模式。我們考慮下列模式：單變量自我迴歸整合移動平均介入模式、時間數列迴歸模式、轉換函數介入模式與指數平滑法，其中轉換函數介入模式中所考慮的投入變數包括育齡婦女總生育率、粗出生率及粗死亡率。我們同時以平均絕對百分誤差 (MAPE) 、根均方百分誤差 (RMSPE) 來評估各模式的預測能力，結果顯示以育齡婦女總生育率為投入變數的轉換函數介入模式最佳，而以粗出生率為投入變數的轉換函數介入模式次之，若以這兩個模式進行未來十年總人口數之預測，並與行政院經建會人力規劃處所作的人口預測中推計值比較，其平均絕對百分誤差分別為0.138%，0.156%，顯示時間數列預測模式有相當佳的預測能力。 / In this thesis, we plan to construct various time series models for the total population in Taiwan. The following time series models are considered: ARIMA intervention model, time series regression model, transfers founction intervention model and exponential smoothing method. The input variable considered in the transfer function intervention model include total fertility rate, crude birth rate and crude death rate. We also compare the prediction performance of these models by using mean absolute percentage error (MAPE) and root mean square percentage error (RNSPE). It turns out that the transfer function intervention model with total fertility rate as input is the best model. While the transfer function intervention model with crude birth rate as input ranks the second best. Finally we forecast the total population of the next ten years by using the above two best models and compare with the middle population projection by Manpower Planning Department in Executive YUAN-Council for Economic Planning and Development. The mean absolute percentage error are 0.138% and 0.165% respectively. This result justifies that the time series model has excellent predictive ability and should be considered for total population prediction.