本研究探討的是在控制公司國家文化背景之下，經理人文化背景對於股利政策之影響。使用Hofstede所提出的個人主義文化價值分數去計算經理人與公司的文化差距，樣本使用S&P1500公司亦有利於控制公司國家文化背景。本文亦探討經理人文化背景形成是否具有不變性，將經理人文化背景分成生涯三階段，分別為幼年階段(出生地)、後青少年階段(就讀大學)以及工作階段(前一份工作)，透過針對生涯三階段的實證分析可以檢測出經理人文化認同形成於哪個階段。 本文發現三個生涯階段形成之經理人與公司的個人主義分數差距對公司股利發放機率皆有顯著正向的影響。本文更進一步發現若當經理人同時為董事長時，上述個人主義分數差距對股利發放之影響便減弱，表示個人主義之影響可能來自於董事會對於高度個人主義經理人之管制。另外也發現經理人與公司的個人主義分數差距與股利增加和開始發放有顯著的正向關係以及與股利降低和停止發放有負向的顯著關係。 本文利用全樣本以及工具變數來進行穩健性分析，全樣本的迴歸分析之下，幼年時期(出生地)與後青少年時期(就讀大學)與原本回歸分析結果相同，但在工作時期(前一份工作)中的顯著性消失。工具變數分析中，當有考慮經理人同時為董事長的變數之下，其幼年時期的結果是穩健的。 ;In this paper, we investigate the impact of the CEO’s individualism on the dividend policy after controlling the individualism of the company. We utilize culture dimensions proposed by Hofstede to calculate the cultural distance between the CEO and the company based on individualism. We use the S&P 1500 index listed companies as our main sample which is advantageous in controlling the individualism of the company. We also investigate whether the time of cultural identity formation matters. We build three different stages of the CEO’s lifetime: the childhood period (the country of birth), the late adolescent period (the country of university), and the working period (the country of the previous job). The empirical findings utilizing the three lifetime stages can tell when the individualism related to dividend policy is formed.We find that the cultural distances based on individualism formed in all three periods have positive impacts on the dividend paying probability. The effect is significantly weakened when the CEO is also the chairman of the board suggesting that our findings could be driven by the board monitoring on CEOs with high individualism. Moreover, we find that the cultural distances based on individualism is positively related to the probability of dividend increase and initiation and negatively related to the probability of dividend decrease and omission. To check the robustness of our results, we use broader samples and instrumental variables. In the broader sample tests, we find the same result in childhood period and late adolescent period, but the effect in working period disappears. We find that the result in the childhood period is robust when the CEO is also a director in the test of instrumental variables.
