由WTO各會員國對原產地證明的重視，可以預期中間財產業的保護與發展，或是自製率的相關產業政策，將持續受到世界各國的重視。而最適自製率如何訂定，自然成為各國政府制定產業政策的首要之務。本論文以最適自製率為研究的主題，研究成果可作為制訂自製率相關產業政策的參考。
若由自製率定義的進一步深究，我們可以發現最終財自製率的高低與中間財的生產、
研發、授權,以及中間財的市場結構是息息相關的。過去有關中間財的文獻，均只考慮單一中間財而已。在本篇論文中，我們考慮多種中間財，而且中間財的種類數目是內生決定的，
這是與過去文獻最大的相異之處。就中間財研發對廠商最適自製率的影響方面，如果假定本國不存在國產中間財市場，而最終財廠商要開發本國所產製的中間財時，唯有投入研發成本。
在不同的要素替代彈性之下，最適自製率解的型式將有所不同。我們發現，要素替代彈性的大小與自製率的高低之間存在反向關係：要素替代彈性高，則最適自製率就較低，廠商可能選擇重點式研發；而當要素的替代彈性較低時，最適自製率則較高，廠商即會選擇全面性研發。當要素的替代彈性為零時，政府對研發的補助居於關鍵地位，因為若補貼額度夠大，
可能使廠商的自製率由最低 0%直接跳至 100%。
隨著中間財技術的創新，可能會出現中間財授權的行為。過去討論授權的文獻，均由技術創新導致生產邊際成本降低的設定出發，這個設定雖然很容易被接受，但也正是其受到限制的地方，不但僅能處理單一中間財的問題，而且必須清楚定義技術創新的幅度。例如,就有線電視的例子,系統業者對消費者提供視訊服務,每一個頻道即是一種中間財,而有線電視系統業者頻道購買的價格與頻道數目等問題,即是傳統的授權文獻所不能處理的。如果我們將廠商技術創新的成果，設定成擁有競爭對手所沒有的特殊中間財，考慮一個最終財產品市場雙佔的模型，由於其中一家廠商擁有特殊的中間財，若授權給對方，一方面使得對手邊際成本降低因而造成自己競爭的劣勢，然而另一方面則可增加授權收入，故而授權定價與種類的決定，與市場規模的大小、要素替代彈性的高低有關。研究發現，當要素的替代彈性越小，
表示競爭對手越需要特殊中間財時，擁有特殊中間財的廠商越不願意授權給對方，均衡的授權種類越少。而市場規模對均衡授權種類的影響，隨著要素替代彈性的大小而有所不同。
隨著生產過程迂迴性的提高，中間財的生產可能會與最終財逐漸分離，並且日漸趨向獨占性競爭的市場結構。如果假定在政府尚未進行自製率管制之前，本國已存在國產中間財市場；而且本國進行自製率管制，對外國中間財的種類數目沒有影響。結合最終財市場與中間財市場之上下游均衡模型，我們發現單純就生產面而言，當政府提高自製率的下限時，
由於中間財種類效果的外部性，國內的中間財種類數目會增加。政府若在最終財廠商原始的最適自製率，亦即無政府管制之下的最適自製率附近做自製率管制的微幅調整，種類效果大於進口替代的負效果，最終財產業的平均成本會下降。而就社會福利極大化而言，
若最終財廠商處於完全競爭的市場，面對直線性的需求函數，存在一個社會最適的自製率，
使得社會福利極大。而當需求規模增加時，最終財的社會最適產量與社會最適自製率均會增加。

From the WTO members'' emphasis on the rule of origin, we can legitimately expect that the protection and development of intermediate-good industries, or the industrial policy related to local content rate, will continue to be respected worldwide. In this dissertation, we discuss the necessary means of determining optimal local content rate, work that we regard as being of value to policy decision-making.
We find that the level of a final-good''s local content rate is relevant to the production, R&D, licensing and market structure of intermediate goods. In the literature, the authors have tended to consider the case of a single intermediate good in isolation. In this dissertation we consider the case of many intermediate goods and regard this variety of intermediate goods as endogenous. The most apparent difference, therefore, between our research and the literature, is that the variety of intermediate goods is endogenously determined.
If we assume that there is no domestic intermediate-good market, any final-good firm with a desire to develop domestic intermediate products would, therefore, have to invest in R&D. We find, as a whole, that the local content rate will be higher when the elasticity of input substitution is lower. Thus, firms will choose very few key intermediate goods on which to invest in R&D if the elasticity of input substitution is lower. In such cases, the government''s subsidy policy in respect of R&D has played an important role. If the amount of subsidy is large enough, the local content rate can perhaps leap directly from 0% to a 100% level.
Due to the continuous innovation in intermediate-good technology, there may exist licensing between different firms. The literature has always assumed that marginal cost is reduced through innovation, however, whilst such an assumption is acceptable, its applicability is extremely limited, as the assumption holds true only in the case of a single intermediate good. Furthermore, there is also a requirement to define the range of innovation. For example, in the cable-TV industry providing program reception services, each channel can be regarded as an intermediate-good. The classic literature is incapable of providing an effective analysis of how we should determine the price charged per channel, and the number of channels licensed to a rival firm.
We believe that a firm''s technological innovation introduces special new intermediate products that its competitors do not have. Assuming that the final-good market is duopoly, as the firm provides special intermediate product licenses to the other firm, it reduces the competitor''s cost, but simultaneously increases its own revenue. We find that the pricing and range of licensing are closely related to the scale of final-good market and elasticity of input substitution. If a firm''s competitor needs these special intermediate products more, its willingness to provide licensing is much lower.
As the process of roundabout production becomes more and more complicated, intermediate-good production becomes disjointed with final-good production, and the intermediate-good market structure, therefore, becomes monopolistically competitive. Assume that domestic intermediate-good market exists prior to government regulation on local content rate, and such regulation does not affect the variety of foreign intermediate products; we discuss the general equilibrium model in respect of these final-good and intermediate-product markets. On the production side, when the government accelerates the rate of regulation, we find that the variety of domestic intermediate products increases because of the externality of intermediate-good''s variety effect. If the new rate of regulation is similar to the original rate, which is the final-good firm''s previous optimal choice, the change in regulation will reduce the final-good firm''s average cost because the variety effect dominates the negative effect that comes from import substitution. On the social welfare side, there exists only one social optimal local content rate, which minimizes the final-good''s average cost if we assume the final-good market is perfectly competitive and that the demand function is linear. When the scale of demand increases, both the social optimal outputs of the final good and the local content rate will also show an increase.