本文?用我國農業部門的總體資?，在生產函?中除?考慮原有之?動與資本投入變?外，?進一步以政府對農業部門公共投資做為非農業部門投入的替代變?，探討我國農業公共投資（公共投資/土地）與農業土地生產?（產出/土地）的互動關係。我們以Akino 與Hayami (1974) 的模型為基礎，針對臺灣地區1952-2001 ?的資?，在考?變?間內生結構性變化（endogenous structural break ）下，檢測農業公共投資與農業土地生產?的長期及短期動態關係。我們的主要發現如下：第一，由共積檢定得出，農業公共投資與農業土地生產?在長期間呈顯著正向關係，農業公共投資的土地生產?彈性為0.555 。第二，在控制內生結構性變化下，農業土地生產?模型的長期均衡關係依舊獲得支持。第三，由弱外生檢定結果指出，長期間農業公共投資與土地生產?互為因果關係，顯示長期間農業部門的成長必須?基於政府對農業部門的公共投資；而在農業部門的成長的過程中，將可?激政府對農業部門的公共投資，?者相互影響。第四，由短期誤差修正模型檢定發現，農業公共投資是體系內失衡發生時的主要調整??。第五，短期間農業土地生產?對

This paper investigates the interrelationship between public investment in agriculture (public investment/land) and agricultural land productivity (output/land) in Taiwan covering the 1952-2001 period. Our main findings are as follows. First, the cointegration test reveals that public investment in agriculture and the productivity of agricultural land both exhibit a significant positive relationship in the long run, where the elasticity of land productivity in relation to public investment in agriculture is 0.555. Next, when controlling for endogenous structural breaks, the long-run relationship for the productivity of the agricultural land model is still supported. Third, the results of the weak exogeneity test indicate that a bi-directional causal relationship exists in the long run between public investment in agriculture and the productivity of land. Fourth, from the short-run error correction model estimation it is found that public investment in agriculture is a major means of adjusting for the disequilibria that occur within the system. Finally, in the short run the unidirectional causal relationship in terms of the productivity of agricultural land on public investment in agriculture is established, otherwise it is not established. From this it can be seen that in the short run the government is unable to reveal the effectiveness of its public investment in agriculture.