小區域死亡率推估之研究
A Simulation Study of Small Area Mortality Projection
真理大學統計與精算系(Department of Statistical and Acturial Sciences, Aletheia University);政治大學統計研究所(Department of Statistics, National Chengchi University);政治大學統計系(Department of Statistics, National Chengchi University)
- 類型
- 研究論文
- 卷期
- 45期
- 出版
- 2012年12月
- 頁碼
- 77–110
- 編號
- NTUEJ00002025
你目前的連線不在校內範圍,所以無法開啟全文。書目與摘要不受限制;回到校內網路後重新整理本頁,全文就會直接顯示。
摘要
臺灣人口結構漸趨老化,由於老年人使用較多醫療等社會資源,人口老化勢必牽動政府政策與資源分配,然而臺灣各縣市的人口老化速度不一(陳政勳與余清祥,2010),因此有必要針對各地方特性發展適當的小區域人口推估方法。小區域推估面臨的問題可歸納為四個方向:「資料品質」、「地區人數」、「資料年數」與「推估年數」,資料品質有賴資料庫與制度的建立,而後三者則與過去資料與推估未來的變異程度有關。本文考量在上述後三個問題的影響下,探討修勻(Graduation)相關方法是否可提高小區域死亡率推估的穩定性。
本文使用屬於隨機推估的區塊拔靴法(Block Bootstrap),以電腦模擬評估推估結果,因為小區域人口數較少,本文也使用Whittaker及標準死亡比(Standard Mortality Ratio)等修勻方法,降低因為地區人數較少引起的震盪。另外,小區域推估通常可用的資料時間較短,未來推估結果的震盪也較大,本文針對需要過去幾年資料,以及未來可推估年數等兩項因素進行研究,希冀結果可提供臺灣各地方政府的推估參考。除了電腦模擬外,本文也以實證分析檢驗修勻的成效,將修勻套用至台南縣、台東縣等縣市層級的死亡率。研究發現,修勻方法可降低小區域死亡率推估的震盪,如有過去十五年資料可獲得較可信的推估結果,而未來推估年數盡量不超過二十年;相對而言,與人口數比較,小區域與大區域的死亡率差異對修勻的影響較為有限。
Abstract
The population size plays a very important role in statistical estimation, and it is difficult to derive a reliable estimation for small areas. The estimation is even more difficult if the geographic and social attributes within the small areas vary widely. However, although the population aging is common phenomenon in the world, the problem is not the same for different countries. The aim of this study is to explore the mortality projection for small areas, with the consideration of the small area’s distinguishing characteristic.
In addition to data quality, the difficulties for small area population and mortality projection can be attributed into three directions: population size, number of base years, and projection horizon. The smoothing methods can be applied to improve the stability and accuracy of small area estimation. In this study, the block bootstrap and smoothing methods are combined to project the mortality of small areas in Taiwan, using the cohort component method.
We found that the smoothing methods can reduce the fluctuation of estimation and projection in general, and the improvement is especially noticeable for areas with smaller population sizes. To obtain a reliable mortality projection for small areas, we suggest using at least fifteen-year of historical data for projection and a projection horizon not more than twenty years. Also, the population size has a bigger influence than the discrepancy of mortality rates between small and large areas.