| dc.contributor | 國貿系 | |
| dc.creator (作者) | Lee, Cheng-Feng;Ho, Tzu-Ping | |
| dc.creator (作者) | 李政峰;何祖平 | zh_TW |
| dc.date (日期) | 2001-09 | |
| dc.date.accessioned | 2-Oct-2015 16:42:33 (UTC+8) | - |
| dc.date.available | 2-Oct-2015 16:42:33 (UTC+8) | - |
| dc.date.issued (上傳時間) | 2-Oct-2015 16:42:33 (UTC+8) | - |
| dc.identifier.uri (URI) | http://nccur.lib.nccu.edu.tw/handle/140.119/78850 | - |
| dc.description.abstract (摘要) | 台灣的總體經濟數列是否具有恆定性，一直是國內實證研究者關心的課題。 然而，常用的單根檢定，在小樣本時，普遍存有型一誤差扭曲與低檢力的問題，進而影響統計推論的可靠性。在過去，許多國內學者曾檢定不同期間的實質GNP數列，結果卻得到不一致的推論。有鑑於此，本文重新檢定台灣的實質GNP數列是否具有單根。除了使用更長的GNP數列(1951-1999)外，研究方法著重樣本訊息的運用，藉由模擬技巧的應用，分別建立在恆定過程與非恆定過程假設下，單根統計量的小樣本分配。透過比較單根檢定值與小樣本分配間的相對位置，判斷實質每人GNP數列最有可能來自於那個分配。實證結果發現：雖然使用目前最長的數列，並充分應用樣本中的訊息，單根檢定的檢力依舊不高，無法明確區分實質每人GNP數列究竟是差分恆定抑或趨勢恆定。數列的長度太短，是可能的原因之一。此外，為考慮石油危機可能引發的結構性改變，文中採用Zivot-Andrews(1992)檢定，將結構改變點的估計視為內生，結果發現，檢力依舊不高，最後結論仍然不變。 | |
| dc.description.abstract (摘要) | Whether real GNP contains a unit root has long been a major concern for applied economists. The literature, however, doucments controversial findings. This paper reexamines the trending behavior of Taiwan`s GNP series. Our approach emphasizes the data`s information content to distinguish between the competing processes. Stationary and non-stationary processes are fitted to the real GNP per capita (1951-1999), the longest among studies following this line. Simulated data are generated, and the small sample distributions of a battery of commonly-used test statistics are computed under each of the two hypotheses. Our simulation exercise indicates that due to low power, all test statistics under study are not very capable of discriminating between the stochastic trend alternative and the deterministic trend one for the series. The series not spanning long enough may constitute one of the major explanations. Furthermore, our conclusion stands well, controlling for the possible structural changes in the series. | |
| dc.format.extent | 156 bytes | - |
| dc.format.mimetype | text/html | - |
| dc.relation (關聯) | Academia Economic Papers, 29(3), 341-364 | |
| dc.subject (關鍵詞) | 單根檢定;實質GNP;隨機性趨勢 | |
| dc.subject (關鍵詞) | Unit root test;Real GNP per capita;Stochasitic trend | |
| dc.title (題名) | 隨機趨勢抑或確定趨勢？再探台灣國民所得數列 | zh_TW |
| dc.title.alternative (其他題名) | STOCHASTIC OR DETERMINISTIC TRENDS? A RE-EXAMINATION OF TAIWAN`S GNP | |
| dc.type (資料類型) | article | en |
