When doing good goes bad

There is a new NBER working paper out on the moral hazard effects of corporate social responsibility.

When Corporate Social Responsibility Backfires: Theory and Evidence from a Natural Field Experiment
by John A. List and Fatemeh Momeni

Abstract:

Corporate Social Responsibility (CSR) has become a cornerstone of modern business practice, developing from a "why" in the 1960s to a "must" today. Early empirical evidence on both the demand and supply sides has largely confirmed CSR's efficacy. This paper combines theory with a large-scale natural field experiment to connect CSR to an important but often neglected behavior: employee misconduct and shirking. Through employing more than 3000 workers, we find that our usage of CSR increases employee misbehavior - 20% more employees act detrimentally toward our firm by shirking on their primary job duty when we introduce CSR. Complementary treatments suggest that "moral licensing" is at work, in that the "doing good" nature of CSR induces workers to misbehave on another dimension that hurts the firm. In this way, our data highlight a potential dark cloud of CSR, and serve to forewarn that such business practices should not be blindly applied.

So "doing good" can result in workers "doing bad".

Incentives matter: shipping file

Research summary: We explore captain-ownership and vessel performance in eighteenth-century transatlantic shipping. Although contingent compensation often aligned incentives between captains and shipowners, one difficult-to-contract hazard was threat of capture during wartime. We exploit variation across time and routes to study the relationship between capture threat and captain-ownership. Vessels were more likely to have captain-owners when undertaking wartime voyages on routes susceptible to privateers. Captain-owned vessels were less readily captured than those with nonowner captains, but more likely to forgo voyage profits to preserve the vessel's safety. These results are consistent with multitask agency, where residual claims to asset value rather than control rights influence captain behavior. This article is among the first to empirically isolate mechanisms distinguishing among major strands of organizational economics regarding asset ownership and performance.

This is the abstract of a new paper in the Strategic Management Journal (Volume 38, Issue 4 April 2017 Pages 854–875) entitled Asset ownership and incentives in early shareholder capitalism: Liverpool shipping in the eighteenth century. The paper is by Brian S. Silverman and Paul Ingram.

The interesting points made by the paper include the obvious, incentives matter, captain-owned ships were less likely to be captured by privateers. The results also suggest that the incentives provided by income rights are weaker than those provided by asset ownership. Of course the protection of assets, the ship, made just be a way of protecting future profits at the expense of current profits.

Managerial control versus performance pay

One of the most obvious problems in any employment relationship is moral hazard. What the boss wants workers to do and what the workers actually do can be two very different things. Two of the most common methods firms utilise to try to deal with this problem are the use of performance pay and managerial control. In the empirical literature the performance pay aspect of this solution has received much greater attention than has the managerial control aspect.

In a recent NBER working paper Kirabo Jackson and Henry Schneider examine how an experiment in managerial control affected revenue at an auto repair firm. When the firm provided detailed checklists to mechanics, and managers monitored their use, revenue was 20 percent higher under the experiment. They compare this effect to that of quasi-experimental increases in mechanic commission rates. The managerial-control effect is equivalent to that of a 10 percent commission increase.

These results suggest that managerial control can be a viable alternative to performance pay for the mitigation of the effects of moral hazard. Furthermore, the managerial-control treatment was larger for mechanics that had higher commission rates, suggesting that in this context managerial control and performance pay are complements. The results also support the theoretical prediction that the optimal incentive contract depends on the level and quality of monitoring.

When they investigated the forces underlying the results, Jackson and Schneider find that the mechanics under the managerial-control treatment increase revenue through doing more repairs on each car and working more hours each week. In contrast, mechanics that received commission increases, increased revenue by substituting away from low-revenue repairs toward high-revenue repairs and getting customers to consent to higher price repairs, with no increase in time on the job or number of repairs conducted. Because this shifting toward more expensive repairs may reflect mechanics exploiting their informational advantage over customers, the result underscore the possibility that pay-for-performance may encourage undesirable worker actions.

Also the behaviour of the mechanics is adversely affected because as they only receive a fraction of the firm's revenue, the additional compensation for mechanics for conducting a more thorough inspection of a car is insufficient to offset the associated effort and time costs. Jackson and Schneider's calculations based on their results indicate that a modest transfer of the profits due to checklist use from firm to mechanic could compensate mechanics for their additional costs and achieve a sizable Pareto improvement.

The Jackson and Schneider study shows that increased managerial control can reduce moral hazard, something that has not often been tested empirically. Evidence is also provided that shows that increased managerial control can generate complementarities with performance pay, most likely in settings with multiple complementary tasks. Given the widespread emphasis on performance pay as an incentivising tool, their results suggest that managerial control may be an additional important tool for designing compensation schemes.