How is employee turnover related to employee retention? A systematic review on two sets of meta-analyses.

Abstract: For years researchers have studied retention and turnover, yet there is still no single solution for the problem of balancing both these concepts in organizations. Turnover is defined as the rupture of ties between an employee and his/her organization, whilst retention is defined as the option of the employee to stay in the organization while not looking out for other job alternatives. These two concepts therefore represent the two sides of a problem that organizations face. The number of relationships between these concepts forms a wicked problem for organizations. The existence of variables in these relationships that change with time, industry and culture makes this problem more abstract and difficult to grasp for organizations. The purpose of this study was to understand how the causes of retention and turnover are related to each other and how both concepts’ consequences are connected between them. Lastly, this study intended to understand how both concepts are related in general. For this purpose, a systematic review of 27 meta-analyses on turnover research and 5 meta-analyses on retention literature was conducted. A quantitative and qualitative analysis of causes and consequences of turnover and retention was performed. Information obtained from the analysis was used to elaborate a diagram of the connections between both concepts in order to provide an overview of the wicked problem. Results show that turnover and retention connections between both concepts’ causes and consequences are opposing to each other. In the relations between causes of both concepts our results show that organizational commitment, job opportunities, job satisfaction and employee performance are the most relevant moderators. In the consequences connections the most relevant relation identified was between employees’ behaviour on behalf of the organization and customer satisfaction. Finally, the overview of the connection between both concepts highlights the wickedness of this problem and the need for relating both concepts in research.