
OpenAI
Meritocracy—the idea that individuals should advance and be rewarded on the basis of their talent and hard work—is one of the most widely celebrated ideals in education, business, and government. It shapes how organizations recruit, evaluate, and promote, promising a fair system where the best rise to the top.
ADVERTISEMENT |
Unfortunately, meritocracy often falls short. Hidden personal biases and entrenched social barriers can undermine fairness in talent management across hiring, performance evaluations, pay decisions, and promotions.
In his new book, The Meritocracy Paradox: Where Talent Management Strategies Go Wrong and How to Fix Them (Columbia University Press, 2025), MIT Sloan professor Emilio J. Castilla offers practical solutions to help organizations develop fairer and more-effective people management practices.
One of the most common—and costly—mistakes Castilla sees in his work with businesses and institutions is what he calls the “best practice trap.” This occurs when leaders look around and mimic what other prominent companies are doing, whether it’s launching a new diversity training program, establishing a merit-driven bonus, or adopting a performance rating system, without first diagnosing what their problem is or reflecting on whether these practices may actually work in their own organizational context.
“If leaders truly want to build merit-based talent management systems, they need to start by looking inward,” Castilla writes.
That process involves “gathering and analyzing data, listening to employees and managers, scrutinizing every step of their recruitment, hiring, performance evaluation, and promotion and pay processes, and asking where bias and other inefficiencies might be creeping in,” according to Castilla.
He outlines the following action items for business leaders and managers who want to make people-management decisions based on data analyses rather than intuition, experience, advice, or guesswork.
Employ a data-driven approach
Making decisions based on data and careful analysis helps leaders take intuition, guesswork, and bias out of the equation. Consider these steps when developing your data-driven strategy.
Identify, develop, and define clear criteria for every major employment decision.
Collect and track key employee features, employment outcomes, and organizational processes systematically over time. (Be aware that in some contexts, collecting certain types of employee information may be illegal, unethical, or otherwise problematic.)
Analyze collected data to identify biases and inefficiencies in your employment decisions.
Decide which intervention to employ to address identified biases and inefficiencies.
Monitor results continuously, reassessing talent needs and updating processes regularly.
De-bias your talent management processes
Bias and other inefficiencies can infiltrate every stage of organizational employment decisions. In his own research, Castilla found that even when employees received the same performance scores, women and Black, Hispanic, and immigrant workers were awarded smaller merit-based pay increases than white men. “It’s not enough to have formal processes in place to guarantee meritocracy in recruiting, hiring, and performance evaluation; bias can easily seep into these systems,” Castilla says.
To de-bias talent management, he recommends that leaders:
• Designate and empower key members of their organization to ensure that biases and social processes don’t lead to unfair treatment.
• Clearly define and communicate the criteria used to advance and reward individuals within their organization.
• Gather and analyze data across recruitment, hiring, evaluation, training, promotion, and retention to identify and address points where bias and inefficiencies become apparent.
Published Sept. 16, 2025, by MIT Sloan School of Management.
Add new comment