• 沒有找到結果。

第四章 結論與建議

第二節 後續研究建議

立 政 治 大 學

N a tio na

l C h engchi U ni ve rs it y

46

本文人口推估考慮生育、死亡兩個面向,礙於資料年數有限,僅能做短年期 (約15~20年) 的人口推估,金碩的死亡率推估研究中建議資料年數至少15年,因 此資料年數和推估年數比最多只能推展到1:1,所得誤差和2:1的結果相當,因 此以1982-2011年、共30年資料推估未來五個縣市人口數時,能確保30年的推估 值,其誤差控制在合理範圍。就生育率來說,由於長年期推估影響年齡層較廣,

以推估N年為例,零歲出生人口將影響之後1~N歲的人口數,對於N歲以上人口 則不具影響,因此,若想更進一步探討長年期誤差,必須針對生育率做更深入的 探討和誤差測試。

本文中,小區域是以縣市層級為單位,為小於國家單位、依地理位置或行政 目的而劃分的區域。通常,小區域人口受到出生、死亡、遷徙的影響,然而本文 中不考慮遷徙,儘管如此,短期人口推估誤差表現仍良好。此外,本文研究方法 也適用小群體規模推估 (小群體係指目標族群,如原住民、弱勢團體、…等等)

,小群體推估不受地域的限制,只需考量出生和死亡。本文關於「出生」的研究

,其應用範圍可以擴大為發生率,舉例來說,若想探討心血管疾病之高危險群,

則有高血量症狀者將被歸入該群體中,視為「發生」。

第二節 後續研究建議

若進一步探討年齡組生育率推估,建議未來三個研究方向:(一) 可以採用 Lee-Carter加權模型,觀察是否結果和區塊拔靴法一致。若兩者結果一致,則表 示使用資料年數的遠近對誤差有決定性的影響。(二) 測試區塊拔靴法的權數,

可以依據MAPE的變化做權數選擇,當MAPE變動越小,該權數設定越佳。(三) 加 入修勻,觀察是否修勻能有效縮小誤差。此外,根據本文結果發現,三個人口推 估面向中,以生育率對短期人口推估的影響最為顯著,建議將生育率分成低、中

、高推估三種情境,測試不同生育率情境對總人口數的影響,並以機率角度詮釋 這三種情境。由於本文電腦模擬採用定常人口假設,建議未來可以使用真實人口 資料 (如世界人口),可排除人口單調遞減假設的影響,也較貼近真實情況。

‧ 國

立 政 治 大 學

N a tio na

l C h engchi U ni ve rs it y

47

本文的人口推估仍侷限在短期人口推估,礙於資料年數不足,無法以交叉驗 證法評估長年期結果,在此提出可能的解決方案,建議更進一步,先以全臺灣資 料,嘗試年數比為30:30的預測 (或較長期間的推估)。以此想法延伸,參考推 估誤差法的概念 (Stoto 1983),假設使用過去30年資料推估未來30年,並且每5 年評估一次,觀察推估結果是否穩定,可判斷推估方法是否存在系統性的誤差。

倘若誤差皆落在合理範圍,則可以假設推估值為真,做更長遠的推估 (這個想法 類似郭孟坤以推估誤差法,提供高、中、低情境機率詮釋)。由於長年期人口推 估有其必要,尤其當已開發國家逐漸邁入高齡化社會的今天,確實掌握未來老年 人口的發展才能預先擬定政策方向和個人退休生活規劃,提前做好防老準備。

本文探討的人口推估主要採用機率推估法,生育率則加入專家意見,而推估 品質倚賴資料品質、資料年數、未來變化和過去趨勢等因素,沒有考量政策、外 在環境 (經濟、醫療等) 的影響。建議未來進行人口推估時,多方參考專家意見

,將更能掌握整體人口的變遷趨勢。另外,資料不足可對缺失的資料進行插補、

嘗試不同的參數 (如生育率下限值)、和納入遷移對人口推估的影響,都可讓推 估結果更接近實際變化,並更進一步評估推估結果對於假設和參數的敏感度。

‧ 國

立 政 治 大 學

N a tio na

l C h engchi U ni ve rs it y

48

參考文獻

中文部分

王信忠、金碩、余清祥 (2012)。小區域死亡率推估之研究。Journal of Population

Studies (TSSCI), 45, 121–154.

余清祥 (1997)。修勻:統計在保險的應用。台北市:雙葉書廊。

金碩與余清祥 (2011)。修勻與小區域人口之研究。碩士論文。

徐茂炫、陳建亨、黃彥豪 (2011)。逾百年臺灣縣市人口興衰之轉折:1987-2010。

人口學刊,43,109–135。

陳政勳與余清祥 (2010)。小區域人口推估研究:臺北市、雲嘉兩縣、澎湖縣的 實證分析。人口學刊,41,153–183。

陳寬政 (1997)。臺灣地區人口出生數量的動態模擬。人口學刊,18,1–18。

黃意萍與余清祥 (2002)。台灣地區生育率推估方法的研究。Journal of Population

Studies (TSSCI), 25, 145–171.

郭孟坤與余清祥 (2008)。電腦模擬、隨機方法與人口推估的實證研究。人口學

刊,36,67–98。

英文部分

Booth H. (2006). Demographic forecasting: 1980 to 2005 in review. International

Journal of Forecasting, 22(3), 547–581.

Brown R. L. (1991). Introduction to the Mathematics of Demography, ACTEXublications, Inc.

Cannan E. (1985). The probability of a cessation of the growth of population in England and Wales during the next century.The Economic Journal, 5(20), 505–515.

‧ 國

立 政 治 大 學

N a tio na

l C h engchi U ni ve rs it y

49

Denton F. T., Feaver C. H., & Spencer B. G. (2005). Time series analysis and stochastic forecasting an econometric study of mortality and life expectancy.

Journal of Population Economics, 18(2), 203–227.

Hall P. (1985). Resampling a coverage pattern. Stochastic Processes Applications,

20(2), 231–246.

Koissi M. C., Shaporo A. F., & Högnäs G. (2006). Evaluation and extending the Lee-Cater model for mortality forecasting: Bootstrap confidence interval.

Insurance: Mathematics and Economics, 38(1), 1–20.

Künsch H. R. (1989). The Jackknife and the Bootstrap for general stationary observations. Annals of Statistics, 17, 1217–1261.

Lawson C. L., Hanson R. J. (1974). Solving least squares problems. New Jersey:

Prentice-Hall, EngleWood Cliffs.

Lee R. D., Carter L. R. (1992). Modeling and forecasting US mortality. Journal of the

American Statistical Association, 87(419), 659–671.

Lee R. D. (2000). The Lee-Carter method for forecasting mortality, with variousextensions and applications. North American Actuarial Journal, 4(1), 80–93.

Lee W. (2003). A partial SMR approach to smoothing age-specific rates. Annals of

Epidemiology, 13(2), 89–99.

Leslie P. H. (1945). On the use of matrices in certain population dynamics.

Biometrika, 33, 183–212.

Leslie P. H. (1948). Some further notes on the use of matrices in population mathematics. Biometrika, 35, 213–245.

Lewis C. D. (1982). Industrial and business forecasting methods : a practical guide to exponential smoothing and curve fitting. London: Butterworth Scientific.

Li N., Lee R., & Tuljapurkar S. (2004). Using the Lee-Carter method to forecast mortality for populations with limited data. International Statistical Review, 72 (1), 19–36.

Myers G. C. (1990). Demography of Aging. Handbook of Aging and the Social Science, Third Edition, 19–44.

‧ 國

立 政 治 大 學

N a tio na

l C h engchi U ni ve rs it y

50

Stoto M. A. (1983). The Accuracy of Population Projections. Journal of the American

Statistical Association ,78, 13-20.

Whelpton P. K. (1936). An empirical method for calculating future population.

Journal of the American Statistical Association, 31, 457–473.

Wilson T., Rees P. (2005). Recent developments in population projection methodology: a review. Population, Space and Place, 19(1), 1–126

Historical data Block bootstrap Lee-Carter

Historical data Block bootstrap Lee-Carter

Historical data Block bootstrap Lee-Carter

Historical data Block bootstrap Lee-Carter

Historical data Block bootstrap Lee-Carter

1992 - 2011 Lee-Carter模型

年齡組推估值

1999 - 2011 Lee-Carter模型

年齡組推估值 總生育率型一推估值 真實值

‧ 國

立 政 治 大 學

N a tio na

l C h engchi U ni ve rs it y

53

相關文件