開發中國家貧窮指數的建構、排名與國家分類之探索

Construction, Ranking and Country Classification of Poverty Index in Developing Countries

研究目的
現有關於建構國家發展指標的研究雖已不少，但從貧窮觀點出發之分析仍相對缺乏。根據2024年全球多向度評量結果，開發中國家之63億人口中約有11億人處於貧困之中。為探討此現象，本研究以貧窮視角建構63個開發中國家的貧窮指數，並進行其貧窮程度之排名與分類。
研究設計／方法／取徑
為瞭解哪些國家最貧窮以及貧窮國家應如何因應等問題，從貧窮觀點探討有其必要性；尤其是針對國際間建構貧窮指數時，因對健康、教育、生活素質等三層面設定相同權重而導致之誤差，本研究改採對10個貧窮變項指標賦予相近的權重之方式，重新建構貧窮指標，實具有學術上的價值。此外，透過對貧窮國家排名與國家分類，除了可進一步地探討落後國家之貧窮問題，更可突顯出人的價值。本研究依據聯合國開發計劃署（United Nations Development Programme, UNDP）（2024）的統計資料，透過主成分分析、相關分析、集群分析、區別分析與多變項變異數分析等方式進行貧窮相關議題之探究。
研究發現或結論
本研究提出之四個假設經過檢定後均為接受，即：一、63個開發中國家之10個貧窮變項值對國家貧窮指數有顯著解釋變異量，且其加權值相當接近，故10個指標應以相等權重建構，而不是如同UNDP，對每個變項給予不同加權。這些國家之中，貧窮程度最低的五個國家為塞爾維亞、亞美尼亞、約旦、泰國、哈薩克；貧窮程度最嚴重的五個國家則為尼日、查德、中非共和國、蒲隆地、衣索比亞，都是非洲國家；二、本研究將63個開發中國家分為高度、中度及低度貧窮三個類別，各有18、13以及32個國家，分類準確度為98.40%；此三類型國家在健康、教育與生活素質及貧窮指數方面都有明顯不同，且與人類發展指數（Human Development Index, HDI）呈現高度顯著負相關，代表貧窮指數穩定性高以及愈貧窮國家的HDI愈低。
研究原創性／價值
本研究之貢獻在於從貧窮觀點探討開發中國家的發展，且不再單以經濟發展瞭解國家發展，而是依據「人的價值」作為國家發展指標，更突顯研究的意義；此外，在建構貧窮指數之各變項時，應注意變項的加權問題。
教育政策建議與應用
本研究發現63個開發中國家的18個高度貧窮國家均集中在非洲，可見非洲國家的貧窮問題相當嚴重，其健康、教育、甚至生活素質各方面均呈現不足的狀態，故應提出脫貧改善計畫，讓這些國家遠離貧窮，以符應聯合國永續發展目標（Sustainable Development Goals）之第一項目標。

Research Purpose
Many existing studies have constructed national development indicators, but there is a lack of research from a poverty perspective. In 2024, a global multi-dimensional assessment of the developing world's 6.3 billion population found that 1.1 billion people were living in poverty.
Research Design/Method/Approach
This study aims to understand which countries are the poorest and how poor countries should respond. Therefore, exploration from a poverty perspective is necessary. International poverty indices often assign equal weights to indicators of health, education, and quality of life, leading to errors. By assigning varying weights to indicators across these dimensions, this study reconstructs a poverty index with greater validity. Ranking and classifying poor countries helps us understand the poverty issues faced by underdeveloped countries and highlight the value of human capital. Using UNDP (2024) statistical data categorized into health, education, and quality of life, this study explores this issue through principal component analysis, correlation analysis, cluster analysis, discriminant analysis, and multivariate variance analysis.
Research Findings or Conclusions
The four research hypotheses proposed in this study were accepted after testing. The results were as follows: 1. The weighted values of the 10 poverty proxy variables for the 63 developing countries were quite close, indicating that the 10 indicators should be constructed with equal weights, rather than the UNDP assigning different weights to each variable. The least impoverished countries were Serbia, Armenia, Jordan, Thailand, and Kazakhstan; the five most impoverished countries were Niger, Chad, the Central African Republic, Burundi, and Ethiopia, all African countries; 18, 13, and 32 countries were classified as high, medium, and low poverty, respectively, with a classification accuracy of 98.40%, showing obvious differences in health, education, quality of life, and poverty indices. 3. Education, health, quality of life index, poverty index, and HDI index were highly significantly negatively correlated, indicating that the poverty index was highly stable and that the poorer the country, the lower the HDI.
Research Originality/Value
This study explored the development of developing countries from a poverty perspective. Rather than focusing solely on economic development to understand national development, it used human values as an indicator of national development, further highlighting the significance of this research. Furthermore, when constructing a poverty index, attention should be paid to the weighting of the variables.
Educational Policy Recommendations and Applications
This study found that the 18 highly impoverished countries were concentrated in Africa. This shows that poverty in African countries is a serious problem, not only in health and education but also in quality of life. Poverty alleviation and improvement plans should be proposed in line with SDG1 to address these multidimensional issues.