Economic agents are motivated to undertake costly actions by the prospect of being rewarded for successes and punished for failures. But what determines what success looks like? This paper endogenizes the criteria for success in an otherwise standard principal-agent model with risk neutrality and limited liability. The set of feasible contracts is constrained by incentive constraints and possibly by a budget constraint. If the principal manipulates the criteria for success only to lower implementation costs, and depending on which type of constraint is more restrictive, the second-best action may be above or below the first-best action. However, given the second-best action, the criteria for success are as stringent as possible. In a class of problems where the principal's payoff depends directly on the criteria for success, the second-best solution features either more stringent criteria for success or a lower action (or both) than the first-best solution.

We use the tools of mechanism design, combined with the theory of risk measures, to analyze a model where a cash-constrained issuer needs to raise capital from a population of different types of risk-averse and budget-constrained investors in order to finance a project with a stochastic return. To achieve that goal, the issuer partitions and sells the project's realized cash flow into several securities, one for each type of risk-averse investor. The optimal partition conforms to the commonly observed practice of tranching with a sequential payment structure (senior debt, junior debt, and equity) where senior claims are paid before the subordinate ones. The holders of more senior/junior tranches are determined by the relative risk appetites of the different types of investors and of the issuer, with the more risk-averse agents holding the more senior tranches.

We consider the problem faced by a durable good monopolist who can allocate a single good at any time, but is uncertain of buyers' values and temporal preferences. We derive conditions under which it is optimal for the monopolist to ignore heterogeneity in buyers' discount factors; for example,
discriminating on discount factor is not optimal when buyers with higher values are also more patient. These conditions also apply when sellers face ambiguity regarding buyers' discount factors. Our results provide a novel justification for temporal nondiscrimination when the seller is incompletely informed about buyers' temporal preferences.

We provide a constructive approach to finding optimal mechanisms for the multi-dimensional multiple-good monopoly problem. Using numerous examples, we illustrate our techniques in settings where agents could have either independently distributed types or correlated types.