中文摘要：

确定最优投资规模是陕西省水利投资决策的关键问题之一.目前已有学者通过实证研究表明,区域性准公共物品在最优供给中存在“拥挤效应”,同时也有学者已从理论上证实,“拥挤效应”会影响准公共物品的最优供给规模.而确定准公共物品最优投资规模的传统方法——道格拉斯函数法,却忽略了其“拥挤效应”属性,因而产生一定局限.那么,陕西省水利设施作为一种具体的区域性准公共物品,其在最优供给中是否也存在“拥挤效应”？拥挤程度如何？如何实证的估计具有拥挤性的水利设施最优投资规模？基于上述思路,本文首先借鉴国外学者对“拥挤效应”的测度方法,建立水利设施拥挤效应测度模型,以验证陕西省水利设施的“拥挤效应”及拥挤程度;其次,将“拥挤效应”的测度方法引入到传统的道格拉斯函数法中,对其进行修正,建立估计拥挤性水利设施最优投资规模的实证模型;最后,运用陕西省1978-2011年时间序列数据,对上述两个模型进行实证估计.结果表明,陕西省水利设施供给存在显著的拥挤效应;若考虑到“拥挤效应”,当人均水利投资约占人均产出的5.74％时,陕西省水利投资水平达到最优;与最优水平相比,陕西省历年的水利投资均低于其最优水平.此外,模型中水利设施供给弹性的估计值（（o）=-0.4418）表明水利设施具有较强的公共物品性质,水利投资的主体应该是以政府为代表的公共部门;同时,人均水利设施供给量对人均补贴的弹性（Φ=0.380 1）为正且数值不小,说明中央政府的财政补助及民间资金对改善陕西省人均水利供给会产生积极影响,因此,陕西省应重视民间资金的引入,积极拓展融资渠道,构建多元化的投融资机制.

英文摘要：

One of the crucial problems about Shaanxi Province＇ s water conservancy investment is that how to decide its optimal scale. Scholars empirically proved that, ＇congestion effect＇ does exist in the optimal supply of regional quasi-public goods, and it is demonstrated theoretically that ＇congestion effect＇ can affect the optimal supply scale of quasi-public goods at the same time. However, the traditional method based on Cobb-Douglas Function has a limitation, because it always neglects the ＇ congestion effect＇ which is a special attribute of quasi-public goods. Then, as a specific regional quasi-public goods, does Shaanxi Province＇ s water conservancy have ＇ congestion effect＇ in its optimal supply.9 How is the congested level and how can we empirically estimate the optimal Scale of congested water conservancy？ Based on the above ideas, this article firstly establishes a model to estimate the existence and its level of the ＇ congestion effect＇ of Shaanxi Province＇ s water conservancy, applying the ＇ congestion effect＇ measuring method created by foreign scholars. Secondly, we introduce the method about measuring ＇ congestion effect＇ into the traditional Cobb-Douglas model, and then modify it to construct a new empirical model which can estimate the optimal investment scale of the congested water conservancy. Finally, we use the annual time series data of Shaanxi Province from 1978 to 2011 , to estimate the above two model. The results show that, the ＇ congestion effect＇ does widely exist in the supply of Shaanxi Province＇ s water conservancy, and the optimal investment level can be achieved when the water conservancy investment accounts for about 5.74% of the GDP per capita, taking ＇ congestion effect＇ into account. Compared to the optimal level, the water conservancy investment in Shaanxi Province has been under the standards over years. In addition, the estimated value of elasticity （Ф = - 0. 441 8） of the water conservancy supply in the model proves that, water