Do employers trust workers too little? An experimental study of trust in the labour market

We conduct a field experiment to investigate employers' trust in workers.  A sample of real entrepreneurs and workers from urban Ghana are respectively assigned to the roles of employers and employees.  Employers have the option to hire (trust) an employee, who can in turn choose whether to exert effort (trustworthiness) in the real-effort task.  By comparing employers' expectations to workers' revealed trustworthiness, we are able to detect potential misperceptions leading to sub-optimal hiring.  We further devise two randomized treatments to test for the existence of expectation bias against specific workers categories and estimate the elasticity of employers' beliefs with respect to new information.  We find that employers significantly underestimate workers' trustworthiness and this reduces their profit.  Employees are aware of employers' sub-optimal trust.  Expectations are largely inelastic with respect to news and negative signals have a strong (downward) effect than positive ones.  Our results suggest that raising employers' expectations would have a strong impact on hiring.