If managers believe, because of formal performance evaluation and compensation systems, that their organization is meritocratic, they are ironically more likely to engage in less meritocratic behaviour. Researcher Emilio J. Castilla of MIT’s Sloan School of management calls this “the paradox of meritocracy.”
Many companies make an effort to promote and compensate their employees based on performance rather than, as in the past, seniority. Specifically, companies will implement merit-based routines and written policies and performance management systems, all with the goal of ensuring that the most effective and successful employees and managers are compensated based on their performance.
In his research, MIT Sloan School of Management professor Emilio J. Castilla demonstrated that well-intentioned, organizational efforts to establish meritocracies through formal evaluation and reward systems may have a surprisingly counterintuitive effect: managers in companies with meritocratic systems are more likely to show demographic bias in their compensation decisions.
Castilla conducted an experiment in which 400 students with management experience, drawn from an MBA program, were tasked with making compensation decisions based on performance evaluations. In the experiment, Castilla manipulated the gender of the employees and whether or not the fictional company for whom they worked (dubbed “ServiceOne”) specifically emphasized meritocratic compensation. (For example, one group of participants read core value statements such as “raises and bonuses are to be based entirely on the performance of the employee,” or, for companies that did not promote meritocracy, “raises and bonuses are to be given at the discretion of the manager.”)
The experiment revealed that participants representing companies with definite meritocratic core values were much more likely to reward men at higher compensation levels than women despite identical performance evaluations — including 12% higher bonuses. (The disconnect between performance and compensation, while still present, was less pronounced in hiring, firing and promotion decisions perhaps because of the visibility of those decisions.)
On the other hand, when “ServiceOne” was not presented in the experiment as a company that embraced meritocratic values, women were likely to receive higher compensation than men.
Why does this paradox of meritocracy exist? Castilla points the finger to a lack of vigilance. Managers assuming that organizational processes will keep bias in check don’t necessarily pay close attention to the impact that their own individual biases, hidden or overt, may have on their decisions. They don’t guard against being influenced by stereotypes. After all, they (maybe subconsciously) think to themselves, given the values of the company, any decision is going to be fair and impartial.
To study the real-world relationship between performance evaluations and compensation decisions, Castilla studied the evaluation and compensation data for 8,900 support staff at a large service organization. The data revealed a bias toward white men. With the same evaluation scores, women received .4% lower levels of compensation and non-white men received .5% lower compensation.
In this company, evaluation and compensation decisions were separated — a common well-intentioned effort to allow supervisors closer to the employees to offer feedback. However, as Castilla’s analysis revealed, this well-intentioned process may have introduced new opportunities for bias.
What is a company supposed to do? Although performance reward systems and policies aimed at curbing biases may have the opposite effect, the answer is not to avoid such systems and policies. Instead, companies must pay special attention to the way such systems and policies are designed and implemented.
Specifically, companies must ensure that their systems are accountable and transparent, with a focus on the processes and criteria of such systems, their outcomes and the intended audiences as follows:
Process accountability: Assign individuals to be responsible for processes, routines and criteria to be used when evaluating and rewarding employees.
Process transparency: Determine which processes, routines and criteria will be visible.
Outcome accountability: Assign individuals to be responsible for which results of pay decisions will be measured for fairness.
Outcome transparency: Decide which compensation decisions and results will be visible.
Audiences accountability: Identify who is accountable for pay processes and outcomes and to whom.
Audiences transparency: Decide who makes pay processes and outcomes visible and to whom.
Castilla implemented these guidelines in the service company whose data he used in his real-world study mentioned above. Three key changes were made to the company’s existing system:
- A cross-functional performance reward committee was appointed to monitor reward decisions, specifically evaluating the fairness of all pay decisions through the use of simple people analytic strategies.
- Senior managers who made compensation decisions had to briefly justify each decision, as a result of the introduction of a new step in the performance-reward process.
- The reward committee had the authority to modify the compensation decisions made by senior managers, especially in situations where demographic disparities (after controlling for performance) emerged.
These simple but important steps led to the elimination of previous demographic pay gaps. The bottom line: stating the core values of a meritocracy, and putting in place the policies and systems are important first steps in creating a meritocracy. However, one final necessary step is to ensure accountability and transparency, thus avoiding the meritocracy paradox of biases caused by a lack of vigilance.
Sign up for free
to receive the Idea of the Week direct to your inboxContact Us
- June 2016
- August 2016