一般文獻對於研發與企業成長的討論,多採取 OLS 迴歸分析法。本文採取追 蹤資料分量迴歸法,除了研究研發與企業成長(即營收成長率、生產力成長率)之間 的關係之外,也進一步研究廠齡對於研發與企業成長二者之間關係的影響。我們使 用 2005 年至 2017 年台灣製造業上市上櫃公司的資料,研究結果表明,研發與企 業成長的關係將受到企業處在不同分量的影響,且廠齡對於研發成效,確實扮演一 個重要的角色: 年輕企業的研發成效(即研發對於企業成長的效果)比老年企業佳, 並且二者之間的研發成效差異,也會隨著分量的上升漸漸擴大。即使將樣本區分為 高科技及非高科技產業,又或是將廠齡透過不同的方式衡量,上述結果皆大致成立。 本文結果顯示,一般文獻採取最小平方法(OLS),不僅無法看出研發對處在不同成 長階段之企業所帶來的不同影響,且將高估或低估年輕或老年企業之研發對企業 成長的影響。
This study draws on the data of Taiwanese manufacturers for the period of 2005 to 2017. The analysis applies a panel quantile regression method to explore the relationship between firm R&D and firm growth in sales and productivity and whether this relationship differs between younger firms and older firms. The empirical results show that firm age does play a crucial role in the relationship between firm R&D and firm growth and the relationship changes with different quantiles. Specifically, we show that the R&D investment by younger firms has a lager effect on firm growth at all quantiles and the difference between the effects of younger and older firms increases when the quantile rises. The above results are still valid for the subsample of high-tech and non- high-tech industries and with different measurements of firm age and productivities. The above results suggest that ordinary least squares method (OLS) not only fails to see the different effects of R&D for firms at different stages of growth, but may also overestimate or underestimate the R&D effects for younger and older firms.