Many firms use subjective performance appraisal systems due to lack of objective performance measures. In these cases, supervisors usually have to rate the performance of their subordinates. Using such systems, it is a well established fact that many supervisors tend to assess the employees too good (leniency bias) and that the appraisals hardly vary across employees of a certain supervisor (centrality bias). We explain these two biases in a model with a supervisor, who has preferences for the utility of her inequality averse subordinates, and discuss determinants of the size of the biases. Extensions of the basic model include the role of supervisor's favoritism of one particular agent and the endogenous effort choice of agents. Whether inequality averse agents exert higher efforts then purely self-oriented ones, depends on the size of effort costs and inequality aversion.