Employee representation on the board of directors and its implications on corporate risk- A quantitative study on diverging stakeholder preferences

Abstract

This study investigates how board-level employee representation (BLER) impacts the risk profile of a firm. More specifically, it investigates whether the binary presence of employees in the board room, or the relative ratio of votes allocated to employees, leads to reduced total, idiosyncratic, or systematic risk. The study extends a small, but growing, body of literature investigating the real economic impacts of the unique form of codetermination that exist in 18 European countries but that is being considered across the world. Hence, the findings have direct political implications. To the authors best knowledge, the study is the first of its kind. The study uses a sample of 3,541 firm-years of Swedish listed firms between 2005 and 2019 to conduct a quantitative analysis. More specifically, the two stage Heckman regression model was used to answer the research question and control for the potential issue of self-selection bias. Based on theoretical arguments from financial risk-return theories, agency theory, and corporate governance mechanisms based on monitoring, information asymmetry and diversity, the study arrived at six testable hypotheses. Firstly, the study hypothesised that the binary presence of BLER would lead to a reduced total (H1) and idiosyncratic (H2) risk but that it would not have an impact on systematic risk (H3). Secondly, the study argued that the marginal effect would be proportional to the ratio of allocated votes. More specifically, it was hypothesised that increased concentration of BLER would lead to lower total (H4) and idiosyncratic (H5) risk, but that it would not impact systematic risk (H6). The study does not find evidence to confirm H1, H2, H4 nor H5. However, it supports H3 and H6. This partially contradicts the most related previous study (Lin et al. (2021)), which has found that employees act as risk-averse bondholders. The authors of this paper provide eight hypotheses that might explain this surprising finding, and argue that the most likely version is that the employee representatives are not given any de facto influence in the board room. Similar arguments have been provided in previous studies, although it is not conclusive nor unanimous. Finally, a word of caution. Although the study rejects the hypothesised BLER-risk relationship, it finds several methodological peculiarities that make the results difficult to interpret objectively. Noteworthy is that polar-opposite results are found when making minor changes to the research design and set of control variables. Hence, the authors suggest several complementary studies that must be carried out before reliable and robust results adequate for political decisions can be derived.