租稅政策是政府常用的工具，透過課稅的對象、方式，稅率的調整或稅式支出等，調整經濟景氣及資源的重分配。本文主要模型是追隨Romer（1987、1990）和Barro＆Sala-i-Martin（1992、1995）而設定，假設對不同的廠商課產出稅，然後搭配不同的稅式支出，共分七種狀況討論，再比較分析、歸納整理共得以下幾點的結論；（1）政府對廠商從事優惠的租稅政策，若增加投資租稅抵減率（或補貼率）和降低稅率將使廠商的R＆D支出增加，且經濟成長率亦會上升。（2）在政府的租稅政策中，不論對那一個廠商課產出稅，只要是對中間財廠商給予ITC或補貼，皆可求解得最適稅率；但若給予最終財廠商ITC或補貼，則最適稅率為零。（3）當政府以經濟成長率為追求目標時，補貼比ITC效果較高。（4）政府所採用租稅政策（不論是ITC或補貼），在α值越接近0.5時，經濟成長率比競爭解的經濟成長率高，此時租稅政策越有效，可視為次佳理論。（5）當政府以產出稅來融通稅式支出，欲達柏拉圖最適境界是一件困難的任務，但若一國的α值非常小時，此時政府所採用的租稅政策將可達成目標。
The R&D production model with intermediate inputs developed by Romer and Barro & Sala-i-Martin is applied in this paper to derive the effect of different tax policy mixes on economic growth. Various combinations of government subsidies, including direct subsidy and investment tax credit, and output taxes on final or intermediate product are treated under a balanced government budget condition. The effect of these tax-subsidy mixes are compared with the original market solution and the Pareto optimality state administered by a social planner. Major findings of these comparisoins are: (1) Preferential tax provisions for R&D investment will have positive effect on business R&D spending and economic growth. This holds in the case of lowering the tax rate, or increases the tax credit or rate of subsidy. (2) Optimal, or growth maximizing tax rate exists if tax break or direct subsidy is provided for intermediate product firms. No matter which firm, final output or intermediate input, is subject to tax. On the contrary, if subsidy is designed for final product producer, the optimal tax rate will be zero. (3) Direct subsidy will give a higher growth rate than investment tax credit. (4) A better than market solution state exists in a tax-intervened world if the output elasticity of intermediate input values about 1/2. (5) The Pareto optimal state can be achieved by the tax tools only if the output elasticity of intermediate input in the production of final goods happens to be very small.
