知識活動與生產力成長－台灣生產農業之實證研究

Abstract

摘要 在農業發展的相關文獻中，農業在供給面的兩種特質，如技術進步以及規模經濟，均是維持農業永續發展的主要原動力。農業生產力的成長可大致區分為要素投入的成長與總要素生產力的成長，在高度工業化的台灣，想要藉由增加要素投入來維持農業的發展是極為困難的，惟有藉由農業總要素生產力的持續成長，台灣的農業才有持續成長與發展的可能。農業總要素生產力的成長大部分來自於技術的進步，而新技術的產生有賴於研究發展等知識活動與人力資本的累積，因此，知識活動對於農業總要素生產力與技術進步的影響程度便是本研究所關注的問題。 本研究蒐集了民國69年至98年間農牧業生產與成本及農業領域研究發展與農業推廣等相關資料，從而計算台灣生產農業的總要素生產力成長情形與技術進步率。為探討農業領域知識累積的時間分配型態及對於生產農業技術進步率的影響程度，本研究首先使用Törnqvist指數計算台灣生產農業歷年總要素生產力的成長率，然後以多項式分配落遲模型估計農業領域公部門研究機構的研發支出落遲期數與落遲型態。最後，本研究利用對偶成本函數建構計量模型，以估算生產農業歷年的技術進步率，並以不同知識活動機構所執行的研發推廣等活動對技術進步率進行迴歸分析，衡量不同研發與知識來源對於技術進步的貢獻程度。 研究結果顯示，在研究期間，台灣生產農業的總要素生產力呈現上升的趨勢。利用二階段法估計多項式分配落遲模型的結果，公部門研究機構研發活動為落後三期、階數二階的落遲型態，且其對於總要素生產力的總效果為正向。總要素生產力的分解結果則顯示生產農業的規模彈性大致呈現逐年遞減的趨勢，生產農業的技術進步率則維持穩定的成長。最後，公部門研究機構的研究發展活動為影響生產農業技術進步率的最主要因素，而其他部門的研究發展與農業推廣活動對於生產農業技術進步率的影響雖然較小，但也有正面的助益。我們由此推論，農業領域的研發與推廣等知識活動是使台灣農業技術持續進步的重要動力，因此也是農業能否永續發展的關鍵因素。

Abstract According to the theory of production, growth of agricultural productivity can be decomposed into two components including input factor growth and total factor productivity growth. Due to increasing input prices over time, it is difficult to maintain the productivity of agriculture by increasing use of input factors in Taiwan. To sustain the development of Taiwan’s agriculture, the growth of total factor productivity is necessary. Since technical improvement is the major source of productivity growth, while technical improvement depends on R&D activities and accumulation of human capital, this study focuses on the impacts of different knowledge activities on the total factor productivity growth of Taiwan’s agriculture. Based on time series production and knowledge activities data from 1980 to 2009, we calculate total factor productivity and technical change rate for Taiwan’s agriculture. To estimate the effects of knowledge accumulation on total factor productivity growth and technical change, the growth rate of agricultural Törnqvist index is constructed. Using polynomial model this study models the lag length and shape of public R&D expenditures of agriculture sector. We then calculate the annual technical change rate of Taiwan’s agriculture, and delineate the tread of technical change through standard regression analysis on various sources of knowledge activities. The empirical results are summarized as follows. First, during the sample period, the total factor productivity of Taiwan’s agriculture revealed a rising trend. Second, estimating the polynomial distributed lag model through a two-stage method suggests that the lag length of public R&D is 3 years and the polynomial degree is 2. Considering the lag structure and polynomial distribution, the effect of public R&D in total is positive. Third, the decomposition of total factor productivity growth indicates production agriculture is characterized as decreasing scale effects and growing technical change. Finally, results from regression analysis demonstrate public R&D activities is the most important factor in improving the technology level of Taiwan’s production agriculture while other minor determinants of agricultural productivity include extension activities and private R&D.