文章詳目資料

地理學報 CSSCIScopusTSSCI

  • 加入收藏
  • 下載文章
篇名 臺灣區域所得探索式時空動態分析,1999-2008 年
卷期 67
並列篇名 Exploratory Space-time Dynamics Analysis of Regional Income in Taiwan, 1999-2008
作者 紀玉臨Yu-Lin Chi周孟嫻Meng-Sian Jhou賴進貴Jinn-Guey Lay
頁次 001-030
關鍵字 區域所得不均探索式空間資料分析空間自相關馬可夫鏈regional income inequalityexploratory spatial data analysisspatial autocorrelationMarkov chainsScopusTSSCI
出刊日期 201212

中文摘要

僅管許多研究顯示區域間存在空間外溢效應,但大多數臺灣區域所得不均的研究忽仍略了此空間面向。本文使用1999 至2008 年綜合所得稅申報核定統計專冊資料,結合傳統不均度指標與空間分析方法,探索臺灣年鄉鎮層級人均綜合所得的時空動態。區域不均分析結果顯示北部-南部、西部-東部、北高-其他、平地-山地、本島-離島、臺灣-金馬等區域所得差距確實存在,尤其是平地-山地的落差最嚴重,且持續擴大。其次,整體的區域貧富差距呈現先惡化,後微幅改善,又急遽惡化的趨勢。探索式空間資料分析則表明區域人均所得有顯著的空間自相關,高所得鄉鎮的空間聚集區分布於三大都會區,而西南部鄉村、山地鄉、澎湖縣與東部區域則為低所得的空間聚集區,且此空間型態十分穩定,表明臺灣區域所得空間極化情形並無明顯變化。最後,馬可夫鏈模型顯示所得移動具有高度黏滯性,低所得鄉鎮要向上轉移、高所得鄉鎮要向下轉移的機率都偏低,且此機率會因鄰近區域的所得高低也有所不同。

英文摘要

It has been acknowledged that spatial spillover effects lead to geographically interdependent regions, yet little research on regional income inequality of Taiwan took spatial effects into consideration. By using traditional inequality indexes and exploratory spatial analysis methods, we explored the space-time dynamics of Taiwan township per capita income over the 1999-2008 period. Regional inequality analysis revealed that the North/South, the West/East, the Taipei and Kaohsiung city/others,the plain/mountain, the main-island/offshore-island, and the Taiwan/Fujian income gaps did exist. Among these regional divisions, the inequality between plain- and mountainous-townships was the most pronounced one. Regarding the temporal trends,the overall regional income inequalities first increased, then decreased slightly, and worsened again sharply. Exploratory spatial data analysis showed that regional per capita incomes were positively spatially autocorrelated. Specifically, three metropolitan areas were identified as spatial clusters of affluent townships. On the contrary, spatial clusters of low income were mainly located around rural areas of the South-West, aboriginal townships, Penghu County and the East part of Taiwan. This spatial pattern was highly stable, implying the spatial polarization of regional per capita income was persistent throughout the study period. Finally, Markov chain models highlighted that the probability of interclass transition was very low, and constrained by its spatial context.

相關文獻