2010年各國人口普查制度之研究

人口普查的主旨在於瞭解一個國家或地區的人口及其相關特性，國家的政策制訂及人力規劃需仰賴正確的調查結果。但隨著社會變遷等諸多因素，即使增加費用也無法提高普查的完訪率，資料的品質也因民眾配合意願而降低。雖然許多國家仍採傳統問卷式普查，為提高資料品質與降低調查成本等因素，部份國家積極發展新的普查方法，預計在2010年人口普查正式實施。本研究整理新的普查方法，包括登記式普查、登記式普查結合抽樣調查、滾動式抽樣調查等方法。其中行政院計畫預計在2010年以登記式普查結合抽樣調查取代傳統的人口普查，效法新加坡、北歐四國與荷蘭的作法，整合公務登記系統以取得普查短表資料，以抽樣調查代替國外長表問卷的設計，獲得較為詳細的教育、生育、居家老人照護等社經議題之統計資料。瞭解登記式普查等新調查方法，有助於2010年台灣普查的進行，尤其是人口普查的目標是常住人口或戶籍人口的基本問題。另外，本研究也將整理美國與法國即將採用的調查方法，包括美國社區調查(American Community Survey)的設計與抽樣方法，用於普查以取代長表問卷的可能，並研究抽樣調查蒐集人口的限制，比較普查及調查蒐集資料的優缺點。

關鍵字

人口普查；登記式普查；美國社區調查；常住人口；戶籍人口

並列摘要

The objective of the population and housing census is to collect demographic information on the population in a nation or an area which will be used as a reference for government planning and policymaking. Because of dramatic changes in the social environment, some problems are generated, such as increases in survey cost, non-response rate, and data demand. Although many countries are still using the Traditional Census method for the 2010 census, some are active in developing new methods to improve the quality of data collected and to decrease the survey cost. Following the examples of Singapore, the Netherlands, and the Nordic countries (Denmark, Finland, Norway, and Sweden), the method for conducting the 2010 Taiwan Census will be different, changing from the traditional census to a registry-based census with sampling survey. This method will integrate the official registry system to acquire the basic demographic characteristics, and collect more detailed information on the social and economic topics, including the aspects of education, fertility, and elder care by using sampling methods. Consequently, this research will be separated into two parts. The first part will introduce and analyze several common census methods used, including the Traditional Census method, Registry-Based Census, Registry-Based Census with Sampling Survey, and the Rolling Census. In addition, the target population of 2010 Taiwan Census is the de jure population, and we shall also discuss its potential problems. In the second part, the research will further examine the survey methods that will be used for the U.S. and French censuses, including the design concept and the sampling method executed by the American Community Survey (ACS) to replace the long-form questionnaire for the 2010 U.S. Census. Finally, we will discuss the restrictions of using sampling to collect data, compare the strengths and weaknesses of census and the different sampling methods, as well as analyze the problems related to the registered population and the permanent residents.