Does Ignoring Heterogeneity in Impacts Distort Project Appraisals? An Experiment for Irrigation in Vietnam
利用越南家庭调查数据，比较了基于均值的灌溉效益估计与考虑异质性的计量方法，发现忽略异质性会高估穷人收益、低估富人收益，并在中高成本下消除净收益。
Could the simplifying assumptions made \n in project appraisal be so far from the truth that the \n expected benefits of public investments are not realized? \n Using data for Vietnam, commonly used estimates of the \n benefits from irrigation investments based on means are \n compared with impacts assessed through an econometric \n modeling of marginal returns that allows for household and \n area heterogeneity using integrated household-level survey \n data. The simpler method performs well in estimating average \n benefits nationally but can be misleading for some regions, \n and, by ignoring heterogeneity, it overestimates gains to \n the poor and underestimates gains to the rich. At moderate \n to high cost levels, ignoring heterogeneity in impacts \n results in enough mistakes to eliminate the net benefits \n from public investment. When irrigating as little as 3 \n percent of Vietnam's non-irrigated land, the savings \n from the more data-intensive method are sufficient to cover \n the full cost of the extra data required, ignoring other \n benefits from that data.