Tuesday, May 3, 2016

Is Meritocracy Dying, Or Was It Really Just a Myth All Along?

An organizational culture that values meritocracy sounds good, doesn’t it? After all, we want to be hired, evaluated, and promoted based on our own merits – as opposed to non-performance-related factors such as our race, our gender, our network, or our political skills. It’s all about getting what we deserve because of our performance.

Scan the web sites of companies and many of them make explicit reference to meritocracy. Goldman Sachs states it is “a meritocracy built on the belief that collaboration, teamwork and integrity create the right environment for our people to deliver the best possible results for our clients.” One of McKinsey & Company’s core values is to “sustain a caring meritocracy.” Similarly, one of Bank of America’s core values is “inclusive meritocracy.” AB InBev, the beer conglomerate, claims that its culture “is built on ownership, informality, candor, transparency and meritocracy. We set ourselves stretch targets and are never completely satisfied with our results.” The newly formed Kraft Heinz Company states on its web site that “we recognize and reward outstanding performance at every level, in the true spirit of meritocracy. In 2014 alone, more than 1,000 employees were promoted as a result of their high performance and value creation.” And there is GE, which has meritocracy as one of its values, which they define as “creating opportunities for the best people from around the world to grow and live their dreams.”

Not only employers but most employees also seem to value meritocracy. According to Ready et al. (2010), this is one of the factors that attracts talent in emerging markets. In my experience, I have also found that young professionals in Asia find the seniority-based system in many Asian firms to be hampering their advancement opportunities, and they want more recognition for their hard work and performance. In fact, when China recruited for its civil service examinations starting around the seventh century, it introduced a merit-based system. The exam was based on classical literature and philosophy, and candidates were selected based on their exam scores. Of course, there were many people from the lower classes who did not have the educational background to compete, and women were excluded from applying. In principle, however, the concept was based on a meritocracy.

Yet there are significant barriers to creating meritocracy in organizations. First, organizations may say they value it but their practices might not necessarily reflect this. Executive coach Marshall Goldsmith (2004) has written about how leaders subtly encourage (and by implication, reward) those who fawn over and suck up to them. According to a recent survey by Right Associates (Lauby, 2012), 44% of employees believe the key to their success lies in “who you know,” and not job performance (39%). Other employee surveys show similar results, despite the prevalence of organizational practices such as performance appraisal systems.  In a 2014 survey of over 356,000 U.S. federal employees, 54% disagreed with the statement that “pay raises depend on how well employees perform their job.” (fedrev report, 2014) In their excellent book, The Meritocracy Myth, McNamee and Miller (2004) point to social capital (“whom you know”) and cultural capital (“what you need to know to fit into the group”) as important non-merit factors that influence life outcomes:
“It helps to have friends in high places, and the higher up one starts in life the greater the probability that one will travel in high-powered social circles. One must also have the cultural wherewithal to be fully accepted within these high-powered social circles. Those who are born into these circles have a nonmerit cultural advantage over those not born into these circles …” (p. 198)

Second, as Castilla and Benard (2010) have argued, emphasizing meritocracy might unintentionally introduce bias and create inequity in the distribution of employee rewards. Their research suggests that “… in contexts in which people are led to feel that they are unbiased, fair, or objective, they are more likely to then behave in biased ways.” (p. 547); they refer to this as the “paradox of meritocracy effect.” Why does this happen? Two mechanisms seem to be operating. One is through the concept of moral credentials. If you believe you have established your credentials as a non-prejudiced person, you are more prone to express prejudiced attitudes.  Similarly, if you believe strongly that your organization values and endorses meritocracy, then you must therefore not be biased and must be making your decisions based solely on merit. The second mechanism is through self-perceived objectivity. If you believe that you are an objective person, then you may not realize that you may be acting on your prejudices. And we all have conscious and unconscious biases. So the paradox here is that if you believe in meritocracy you might be more likely to behave in biased ways.

Third, we are subject to the fundamental attribution error, which is the tendency to overattribute other people’s behavior to internal rather than to external causes, and one’s own success to internal rather than external causes. Those who have made it and who are successful believe that it is due to their hard work and because they got there through their merits. They also believe that internal factors such as aptitude or motivation are the causes for others not succeeding, rather than external circumstances.

A fourth barrier is around our unconscious biases, especially related to (but not limited to) race and gender. As an example, Samson (2013) conducted a survey-based experiment in which half of the sample of White respondents in California were asked about the importance of GPA (a merit-based variable) as a factor for admission to the University of California. The other half received the following information prior to the question: Under current admissions procedures in the University of California system, Asians make up almost 40% of the student body (or 2 out of every 5 students), while they are only 12% of the California population (p. 240). His results showed that individuals’ acceptance of merit-based admission varies as a function of their perception of “group threat” and not just perceptions of fairness. In other words, when group threat is primed (as it was in half the subjects), these White subjects tended to decrease the importance of a merit-based variable like GPA as a factor in admissions. He concludes that commitment to meritocracy may be based not solely on principle but varies “… depending on the outgroups under consideration and the extent of the group threat they pose to Whites.” (p. 253)  

In an interesting set of studies, Uhlmann and Cohen (2005) showed that individuals use different criteria to define merit that fits their own preconceptions and biases. For example, in one study, they asked participants to rate the strengths of male and female applicants for the job of police chief. Applicants’ credentials and gender were manipulated. Credentials were either strong on being streetwise (e.g., working in tough neighborhoods, getting along with fellow officers, poorly educated) or on being educated (e.g., experienced in administrated, well educated, did not get along so well with fellow officers). The participants then rated the importance of several hiring criteria, and whether the applicant should be hired. What they found was that educated characteristics were more important when the male applicant had them than when he did not. Educated characteristics were less important when the female applicant had them. In fact, male and female participants constructed criteria favorable to the male applicant. Because the job of a police chief is traditionally male, the participants defined merit in a way that favored male over female applicants.    

As reported in the Wall Street Journal (Silverman, 2015), the Clayman Institute for Gender Research at Stanford has been analyzing hundreds of performance reviews in several technology companies. So far, researchers have found that women received 2.5 times the amount of feedback men did about aggressive communication styles (e.g., suggesting that they pipe down and be less aggressive), while men’s reviews contained twice as many words related to assertiveness, independence, and self-confidence (e.g., “drive,” “transform,” “tackle”). In addition:

“… women’s reviews had more than twice the references to team accomplishments, rather than individual achievements … while men … received three times as much feedback linked to a specific business outcome and twice the number of references to their technical expertise.”

Fifth, the criteria for determining merit can vary by company and by culture. Being promoted in some Asian companies means that you have to be able to speak English very fairly well. Many companies look at factors other than performance to determine who gets hired, rewarded and promoted, and arguably some of these factors are more subjective than others, as well as being weighted differently by different raters. For example, in identifying those who might be the future leaders of the company, executives might consider performance, but often look at other factors, such as interpersonal skills, that elusive quality called executive presence, and perhaps even the advocacy of other executives who might have mentored these individuals. Aren’t some of these “non-merit” factors? McNamee and Miller (2004) give a detailed example of their own experience selecting a new faculty member for their department, and they conclude:
“… we call into question the presumption that people know merit when they see it. … it is a cardinal principle in meritocracy that the ‘most’ qualified or ‘best’ person should be hired for the job … However, we argue that it is often difficult or impossible to ‘know’ who the ‘best’ is.” (p. 43)

What can organizations do? First, and the most basic, is to make sure their practices are consistent with their statements about merit. Do their performance and reward systems clearly reinforce excellence and high performance? They can conduct periodic surveys or pulse checks to determine what employees’ perceptions are, instead of simply relying on making pronouncements. Second, make these practices and processes more transparent by defining more clearly what excellent performance is, and why it matters. For example, if behaviors in addition to performance are used as criteria for rewards and promotions, make these explicit, and build in safeguards to minimize subjectivity.

Third, hold managers accountable. Have processes in place to make sure that managers evaluate and reward employees based primarily on performance and other merit-based factors. There is a lot of research evidence from studies of small groups that meritocratic hierarchies (for example, when individuals in groups are given higher rank because of their task expertise) predict group success (Anderson and Brown, 2010).

Fourth, raise awareness of unconscious biases. According to the Wall Street Journal (Lublin, 2014), as many as 20% of large corporations with diversity programs (including Chubb, Genentech, Google, Price Waterhouse Coopers, Roche and T. Rowe Price) now provide training on unconscious bias. Other companies (e.g., Microsoft) provide guidelines to managers before they write up their performance reviews to remind them about gender bias.  There is some evidence that such training might be helpful. Of course efforts to consider diversity in hiring and promotion decisions should continue. Especially important is making sure that there is diversity in the pool of candidates being considered. Most White senior executives when considering successors for their positions or those within their team might favor first those who they perceive to be like them (the “similar-to-me” bias). With a more rigorous screening process, organizations can make sure there are candidates who may be just as qualified but who may not be on their radar screen – and some of them might be female and people of color.

Fifth, as much as possible, define criteria for merit up front and get agreement from evaluators on what the indicators for these criteria are. This will reduce individuals’ tendencies to emphasize qualifications that tend to fit their biases. When working with organizations on evaluating their future leaders, make it a practice, as my colleagues and I do, to gain agreement with the executive team on what success looks like in concrete terms.

It is clearly not enough for organizations to simply state that they have a meritocratic culture. In fact, the reality is that it is very difficult for any organization to achieve this ideal, although it is well worth the effort. The current debates about wage and income inequality touch on the difficulties of achieving meritocracy in societies. However, companies that are serious in attracting and retaining talented employees need to continue to make serious efforts to eliminate these barriers.

Anderson, C. and Brown, C. (2010). The Functions and Dysfunctions of Hierarchy. In A. P. Brief and B. Staw (Eds.), Research in Organizational Behavior, 30: 55-89. New York: Elsevier.

Castilla, E. and Benard, S. (2010). The Paradox of Meritocracy in Organizations. Administrative Science Quarterly, 55: 543-576.

Kaplan, S. (2015). Meritocracy: From Myth to Reality. Rotman Magazine, Spring, 49-53.

Lublin, J. (2014). Bringing Hidden Biases into the Light. Wall Street Journal, January 9:  http://www.wsj.com/articles/SB10001424052702303754404579308562690896896

Lauby, S. (2012). For Your Career, It’s Not What You Know – It’s Who You Know. http://blog.shrm.org/blog/for-your-career-its-not-what-you-know-its-who-you-know

Samson, F. (2013). Multiple Group Threat and Malleable White Attitudes Towards Academic Merit. Du Bois Review, 10(1): 233-360.

Silverman, R. (2015). Managers: Watch Your Language. Wall Street Journal, September 30 2015.

Uhlmann, E. and Cohen, G. (2005). Constructed Criteria: Redefining Merit to Justify Discrimination. Psychological Science, 16 (6): 474-480.