| 摘要: | 國際援助的有效性日益受到重視，台灣雖然國際空間有限，仍持續進行對外援助，並積極遵循近年來國際組織與他國外援機構使用數據及更嚴謹科學方法檢視援助計畫成效的做法。本研究計畫檢視財團法人國際合作發展基金會（國合會）「加強中美洲農牧保健組織轄區柑橘黃龍病防治及落實病蟲害綜合管理計畫」之成效，針對台灣2012至2017年對於貝里斯、宏都拉斯、瓜地馬拉、尼加拉瓜、薩爾瓦多五國的援助進行評估。在發展經濟學文獻中，植物病蟲害防治技術屬於農業技術採用的一種，關於技術採用的文獻雖然豐富，卻少見直接連結到家戶與個人福祉影響的研究結果。因此本研究預計使用國合會在貝里斯、宏都拉斯、瓜地馬拉、尼加拉瓜四國收集的初級資料，以及計畫主持人隨國合會至上述四國加上薩爾瓦多的移地研究，從區域、業者、家戶、到個人層級，探討黃龍病防治計畫對於區域內所得分配、農業收益、家戶與個人福祉的影響，並進一步探究同一區域的業者間是否存在外溢效果，以及家戶內是否存在異質效果。實證上除了以最小平方法與三重差分對於受益戶、援助區域中的非受益戶、非援助區域中的純控制組進行比較，亦將使用傾向分數配對法、家庭固定效果與空間不連續迴歸處理內生性問題。
The effectiveness of foreign aid has received more attention and scrutiny recently. Although Taiwan faces challenges of international relations, Taiwan constantly gives foreign aids and actively follows the practices of international development organizations and foreign aid agencies of other countries: using data and more rigorous scientific methods to examine the impact of foreign aid projects. This research proposal attempts to evaluate the impact of TaiwanICDF's 2012-2017 Strengthening the Control of Huanglongbing (HLB) and the Implementation of Integrated Pest Management (IPM) in Citrus Project with the focus of five participating countries: Belize, Honduras, Guatemala, Nicaragua, and El Salvador. In the literature of development economics, pest control belongs to the discussion of agricultural technology adoption. Despite the abundant literature of technology adoption, research about the impact on household and individual well-being is rare. Therefore, this research proposal will evaluate the impact of the HLB Project on regional income distribution, agricultural output, and well-beings of households and individuals. Potential spillover effects across farms and heterogeneous effects within households will also be explored. Data will come from TaiwanICDF's survey data collected in Belize, Honduras, Guatemala, Nicaragua, and the information collected by the Principal Investigator (PI) with TaiwanICDF staff during the field study in the above countries and El Salvador. Empirically, in addition to using OLS and difference-in-difference-in-differences (DDD) to compare participants, non-participants in beneficiary areas, and samples in non-beneficiary areas, the PI will Propensity Score Matching (PSM), family fixed effects (FE) and spatial regression discontinuity (RD) for identification. |
