When Paying More Doesn’t Pay Off: A Critique of Incentives

Tomasz Obloj, assistant professor at HEC Paris (Strategy and Business Policy Department) says most organizations do not anticipate the actual consequences of the incentives they deploy.

As per his research, there are two important problems created by incentives: the incentives life cycle, or ‘cobra effect’, and the incentive competitive effect, or ‘blood donor dilemma’. What are these you ask? Read on or hit the video towards the end of this page for a more in-depth look at these phenomena. Here is the cobra effect in more detail. 

Many years ago, the British governor of colonial Delhi, in order to free the city from a plague of snakes, decided to put in place an incentive regime introducing a bounty on cobra skins. The bounty was quite high as cobras are tricky to catch … unless they need not be caught. It turned out that it was now a sound business idea to start farming snakes instead of risking life catching them. And so, all of a sudden, the number of bounty claims increased disproportionately to the actual gravity of the problem. The local authority realized the problem and responded rationally – they abandoned the incentive scheme. And so the cobras were released from the farms into the city!

While incentives usually lead to increasing the measured performance, they do not necessarily lead to increasing the true underlying objective

Other examples of such perverse incentive consequences abound. A former star quarterback of New York Jets, Ken O’Brien was said to have stopped passing the ball in situations where he should have, as soon as his team included a financial penalty for interceptions in the contract.

The Sears Holding Corporations abolished the commission plan in its auto centers once its employees were discovered to have been notoriously leading customers to authorize unnecessary car part replacements. The initial bonus plan was based on a profit-sharing rule where mechanics benefited financially from every new component installed. 

Many of the observers of the recent financial crises attribute its origins, at least in part, to extremely powerful short-term incentives granted to investment bankers. Despite these facts, a vast majority of the companies around the globe rely on some form of pay-for-performance in their employment contracts.

So what do we know about incentives that could prevent such disastrous consequences? First, we know that they work. Incentives are a very powerful tool affecting our actions. Past several decades of research in social sciences as well as casual observations of the world show unambiguously that both financial and non-financial incentives change behavior, reinforce performance, and induce more effort. This is an undisputed fact and this is of course where the power of incentives lies.

However, we also know a second thing, that when incentives are concerned, one should be careful with what one wishes for. This is because while incentives usually lead to increasing the measured performance, they do not necessarily lead to increasing the true underlying objective. The number of cobras that were traded for bounty increased – that was what was measured. This however did not decrease the potential danger from snake plague in the city. Similarly, the number of interceptions that Ken O’Brien threw also decreased subsequently to a change in his contract. This however came at a cost of potential successful passes that were never made. Finally, bank employees issued a record number and volume of mortgage loans in the early XXth century. The problem was however, what was not measured by their incentives the long-term quality of these loans.

Despite these unintended consequences of financial incentives, firms may often be tempted to accept the costs in a hope that the productivity increases will outweigh them. This may be true in some cases but there is evidence that should caution us even more against using powerful compensation mechanisms that are based on imperfect proxies for effort

Despite these unintended consequences of financial incentives, firms may often be tempted to accept the costs in a hope that the productivity increases will outweigh them. While this may be true in some cases, there are two more recent pieces of evidence that should caution us even more against using powerful compensation mechanisms that are based on imperfect proxies for effort. 

First, the efficiency of incentive systems evolves over time. While the productive effects are most pronounced early on, the adverse consequences manifest themselves more and more as employees learn the loopholes in the system and figure out how to game it. I call this process the incentives lifecycles. What it means is that, just like with products, incentives need to be redesigned from time to time in order to reset the adverse learning clock. The exact frequency will of course be specific to a given industry and firm but my research shows that a period of 2 to 3 years usually constitutes an optimal frequency of change.

Second, the negative consequences of incentives are inherently linked to the ability of employees in a firm. In fact, the smarter the employees, the more exposed is a firm to the dark side of incentive systems. It is not therefore a coincidence that a best-selling chronicle of the Enron story of dubious financial engineering, written by Bethany McLean and Peter Elkind, was entitled “The Smartest Guys in the Room.” This problem is especially vexing because the productive effects (bright side) of incentives also increase with the ability of employees.

In economic terms, the power of incentives and ability should be complements: the benefit to one increases with the magnitude of the other. However, if the costs are also exacerbated by ability, high human capital firms should be especially careful to include in their incentive design the long-term consequences of employee actions. After all, Wall Street employed and still does some of the brightest talent. (Images courtesy healthandfitness.com & salesforcesearch.com)