The Degree of Internationalization and Firm Productivity: Empirical Evidence from Large Multinationals
利用人工神经网络半参数模型，研究大型非金融企业1992-2019年国际化程度与生产率的关系，发现倒U型关系，且国际化对技术创新有正向非线性影响。
Abstract This paper examines whether the S‐shaped and M‐shaped hypotheses explain the internalization firm's productivity relationship. The internationalization–performance (I–P) literature uses accounting‐based performance indicators in order to examine such a relationship. In contrast to the mainstream literature, productivity and its components (efficiency and technical change) are used as a firm's performance measures. Utilizing a semi‐parametric model based on artificial neural network techniques, accounting for potential heterogeneity, firms’ productivity, efficiency and technical change levels are estimated. The innovative methodological framework is applied in a sample of large, experienced non‐financial firms over the period 1992–2019. The empirical evidence suggests that firms’ internationalization in relation to their productivity and efficiency levels exhibits an inverted U‐shaped relationship. This finding corresponds to the last two stages of the S‐shaped and M‐shaped hypotheses. Furthermore, the evidence suggests that internationalization has a positive nonlinear effect on firms’ innovation capacity (technical change). Overall, the empirical findings from data‐driven techniques applied, support the view that the effect of internationalization on firms’ productivity levels is asymmetric.