In many organizations the measurement of job performance can not rely on easily quantifiable information. In such cases, supervising managers often use subjective performance evaluations. We use laboratory experiments to study whether the way employees are assigned to a manager affects managers' and co-employees' subjective evaluations of employees. Employees can either be hired by the manager, explicitly not hired by him and nevertheless assigned to him or exogenously assigned to him. We present data from four different treatments. For all four treatments we find escalation bias by managers. Managers exhibit a positive bias towards those employees they have hired or a negative one towards those they have explicitly not hired. For three treatments we find that managers' and employees' biases are connected. Exogenously assigned employees are biased in favor of employees hired by the manager and against those explicitly not hired.

Full bibliographic details must be given when referring to, or quoting from full items including the author's name, the title of the work, publication details where relevant (place, publisher, date), pagination, and for theses or dissertations the awarding institution, the degree type awarded, and the date of the award.

Publicly provided goods often create differential payoffs due to timely or spatial distances of group members. We design and test a provision mechanism which utilizes rank competition to mitigate free‐riding in impure public goods. In our Rank‐Order Voluntary Contribution Mechanism (Rank‐Order‐VCM) group members compete via observable contributions for a larger share of the public good; high contributors receive preferential access (a larger share), while low contributors receive restricted access (a lower share). In a laboratory experiment, Rank‐Order‐VCM elicits median contributions equal to the full endowment throughout the finitely played games with constant groups. In the control treatment, with randomly assigned ranks, the contributions are significantly lower and decline over time. We thus provide evidence of rank competition, in situations where discriminatory access to public goods is possible, being efficiency enhancing. (JEL C91, H41)

In a circular neighborhood of eight, each member contributes repeatedly to two asymmetric (i.e. with different freeriding incentives) local public goods, one with the left and one with the right neighbor. All two-person public good games provide only local feedback information and are structurally independent in spite of their overlapping player sets. Here heterogeneity across neighbors is induced by two randomly selected members, named "Bad" Apples, who are either less productive or excluded from periodic information feedback about their payoffs and neighbors' contributions. Although the presence of both "Bad" Apple types leads to the neighborhood, as a whole, evolving less cooperatively, the way in which it spreads is quite different. While less productive "Bad" Apples directly initiate the spoiling of the basket due to their low contributions, "Bad" Apples excluded from periodic information are exploited by their neighbors. Furthermore, we find evidence that "Bad" Apples' positioning affects contributions in the neighborhood.

No abstract