Ability, Adverse Learning and Agency Costs: Evidence from Retail Banking

The literature on incentive contracting suggests that the optimal performance pay contract depends on a tradeoff between productivity and agency costs. The effect of employee ability on this tradeoff is theoretically ambiguous, as the employee’s private gains (through more sophisticated gaming responses) may exceed the employer’s productive benefits. Similarly, employees’ “adverse learning” about how to game their incentives may outweigh their productive learning. Existing research has not examined these possible perverse effects. We observe branch managers of a large retail bank following the introduction of a new incentive plan. We use a novel empirical strategy to estimate the profits the bank loses through managers’ manipulation of loan sizes and interest rates, and find that these agency costs are between three and twelve percent of profits on average. Managers’ formal education (“book smarts”) has no impact on agency costs, but their ability to infer undisclosed information about the incentive plan (“street smarts”) does. More-able managers in the latter sense cost the bank an extra two percent of profits. Finally, agency costs are increasing over time, suggesting that adverse learning dominates productive learning. We find suggestive, but inconclusive, evidence that higher levels of “street smarts”are associated with a higher rate of adverse learning.