Many large firms award year-end bonuses based on the total profits earned by a company or a division. But, with the exception of a firm’s top executives, such bonus plans are a poor incentive for rank-and-file employees, says Michael Gibbs, clinical professor of economics at Chicago Booth.
In all but very small companies, each individual employee is unlikely to have a noticeable effect on the entire company’s profits. “The signal-to-noise ratio is nearly zero, so such an approach is tantamount to using a lottery ticket as the performance measure,” writes Gibbs in a 2012 study.
Plans like these are probably used because they are easy to administer, but that’s not a good justification if they don’t motivate. Instead, an employee’s evaluation should reflect the effects of his or her actions on the company. Performance reviews are the most important component of a good incentive plan, says Gibbs. But choosing the right mix of numeric and subjective measures is tough—and important.
Problems with numbers
Numeric goals have an advantage: employees know how they will be evaluated and how close they are to achieving objectives. The goals leave little room for argument and are, in theory, beyond manager discretion. However, no numeric measure is perfect.
One potential flaw is that numeric measures ignore intangible activities such as service and innovation. A manager whose bonus is based on annual profits might avoid spending money on plant maintenance or research and development, for example. Another flaw is that company performance measures such as stock prices can be too broad to motivate employees.
While incentive pay shouldn’t depend on uncontrollable risks, companies ought to reward employees for uncertainties they can manage, says Canice Prendergast, W. Allen Wallis Professor of Economics and Booth Faculty Fellow at Chicago Booth. Imagine a firm with large-scale construction projects around the world, Prendergast says. The company has more experience in, say, Canada, than in Armenia. “If one were to make predictions about compensation using the standard trade-off between risk and incentives, one would expect to see more pay for performance for the project manager in Canada than in Armenia, since the manager’s performance can be measured more precisely in Canada,” writes Prendergast.
If the Armenian manager’s performance and pay are tied to the profitability of the project, the uncertainty in her pay is beneficial for her and the company. By tying pay to the performance of the project, the company delegates decision making to the Armenian manager, which compensates for the firm’s lack of information about the country and pushes the project manager to figure out the best way to perform the job.
The information advantage that employees have over their supervisors can sometimes be used to manipulate performance measures in damaging ways. Consider a hospital that measures surgeons’ performance solely by the number of deaths on their watch, says Prendergast. Surgeons could improve their evaluation by operating only on patients with simple medical problems, avoiding more complex cases.
The importance of judgment
Subjective evaluation by supervisors can address the shortcomings of numeric measures, Gibbs says. When numeric measures focus employees on one goal, a second, subjective bonus can make employees pay more attention to other objectives that may be difficult to quantify, like managing controllable risks. If a plant manager’s bonus depends on profits alone, he might postpone maintaining equipment. A supervisor can motivate the plant manager to consider the longer-run profitability of the machines if part of the bonus is discretionary.
A subjective review can also help deter manipulation if employees know their supervisors will assess them beyond numeric targets. If a supervisor suspects that an employee has manipulated a performance measure, the employee will be penalized accordingly. Foreseeing this may deter such manipulation in the first place.
In a survey of 250 US car dealerships, Gibbs and his coauthors find empirical support for these ideas. Discretionary bonuses mitigated the flaws inherent in numeric-based bonuses, helped motivate managers to make long-term investments, rewarded cooperation across departments, and kept sales managers motivated during events such as an economic downturn, when sales targets were tough to meet. In other words, they allowed owners to avoid punishing employees too much for bad luck that was not within their control. Finally the research team found that discretionary bonuses were more likely to be used if numeric measures were reported to be easier to manipulate—evidence that subjectivity is used to punish and deter manipulation.
Granted, subjective evaluation isn’t perfect. “Subjectivity opens the door to favoritism,” write Prendergast and Robert H. Topel, Isidore Brown and Gladys J. Brown Distinguished Service Professor in Urban and Labor Economics at Chicago Booth, in a research paper on favoritism in organizations. Discretion allows supervisors to play favorites with their staff. Knowing that bosses can be bribed encourages employees to spend time currying favor with bosses, ignoring more productive tasks. It’s also harder for companies to tell who truly deserves a promotion or reward if supervisors tend to favor certain employees. A good performance rating could be due to either favoritism or a genuinely good performance.
Even a hint of favoritism in the workplace could make it hard for employees to trust their supervisors. “If subordinates do not trust their evaluators to make informed and unbiased assessments, then the result could be employee frustration, demotivation, and turnover,” write Gibbs and his coauthors.
In the study on car dealerships, managers who had been working at a dealership for a long time, which suggests that they were more likely to trust their supervisor to give them a fair evaluation, were happier with the discretionary bonuses they received. Rewards based on subjective reviews also increased sales productivity and net profit per employee among sales managers with more tenure at a dealership. Greater trust makes incentive plans more effective.
Empirical studies of subjective evaluation are rare, because it is difficult to quantify the concepts. The paper by Gibbs and coauthors that focused on this topic was awarded the Notable Contribution to Management Accounting Research award by the American Accounting Association.
This article originally appeared alongside Are CEOs overpaid? The case against.
Michael Gibbs, “Design and Implementation of Pay for Performance,” Oxford Handbook of Managerial Economics, Oxford University Press: July 2013.
__________, “Designing Incentive Plans: New Insights from Academic Research,” World at Work Journal, Fourth Quarter 2012.
Michael Gibbs, Kenneth Merchant, Mark Vargus, and Wim Van der Stede, “Determinants and Effects of Subjectivity in Incentives,” Accounting Review, April 2004.
Canice Prendergast, “The Provision of Incentives in Firms.” Journal of Economic Literature, March 1999.
__________, “The Tenuous Trade-off between Risk and Incentives,” Journal of Political Economy, October 2002.
Canice Prendergast and Robert H. Topel. “Favoritism in Organizations.” Journal of Political Economy, October 1996.