Exaggerated Competence

As human beings, we tend to overestimate our abilities and underestimate our faults. In one survey, 84 percent of middle managers and 97 percent of executives claimed to be among the top 10 percent of performers in their organization.¹ So common is the habit of overrating one’s abilities that it has a name: the better-than-average effect. One oft-cited meta study found that the correlation between self-assessed and actual performance was just 0.29 and, in the case of managerial performance, a paltry 0.04.²

While the inclination to self-aggrandizement is universal, it’s particularly pronounced at the top. Here’s why.

First, highly confident people tend to have an advantage in competing for power. Research shows that in judging the competence of others, we’re heavily influenced by bluster. The more confident someone appears, the more likely we are to believe they’re genuinely capable, whether or not that’s true. Genuine competence is often hard to assess, so instead we gauge an individual’s self-confidence. Working with colleagues at the University of California, professor Cameron Anderson conducted six studies on overconfidence and social status. The research strongly confirmed the proposition that “overconfident individuals [are] perceived as more competent by others.”³ The implication: it’s often the most confident people, not the most competent, who get to the top. Stated more bluntly, the gap between self-perception and reality is likely to be greatest where the air is thinnest. In case you had any doubt, it really is possible to bullshit your way to the top.

Second, in a formal hierarchy, power relationships are highly asymmetric. Managers have a lot more control over their subordinates than the reverse. This makes it risky to question a superior’s competence. Stick a pin in your boss’s overinflated ego and it’s your career that will go “pop!” Power differentials encourage acquiescence, which leaders often mistake for agreement. It’s more gratifying to believe that a sea of nodding heads betokens assent than to entertain the hypothesis that one’s subordinates are merely buying career insurance. In the presence of the powerful, discomforting facts get ignored, contrary opinions go unexpressed, and doubts about executive competence are raised only in hallway whispers.

There’s a third reason hierarchy promulgates unrealistic assumptions about executive competence. Among those who subscribe to a top-down view of authority, there’s a common belief that “big” issues are the sole preserve of “big” leaders. While it’s true that the senior leaders are ultimately accountable for strategy, it doesn’t follow that they’re the best ones to create it. There’s only so much wisdom and experience within the executive team—and it’s often not enough. Yet senior leaders are often reluctant to crowdsource strategy. After all, how can they justify their generous pay packets if they’re not the ones plotting the future and making the “big calls”?

That’s the problem with formal hierarchy: leaders are expected to make crucially important decisions on precisely the sort of complex and ambiguous issues that exceed the cognitive limits of any small group of individuals. As we argued in chapter 2, hierarchy asks too much of too few. Unfortunately, executives often believe they’re up to the task.

Take the case of Jeff Immelt, the chairman and CEO of General Electric from 2001 to 2017. A few of Immelt’s decisions, like selling GE’s plastic business, were widely praised. Unfortunately, these moves were not enough to offset a plethora of questionable bets—like bulking up GE Capital just before the financial crisis, overpaying for French power company Alstom, and sinking $93 billion into stock buybacks while loading up on debt. During Immelt’s tenure, GE’s stock rose by a scant 27 percent, compared with 183 percent for the Dow Jones Industrial Average. In interactions with outsiders, Immelt came across as smart, charming, and eager to learn, yet internally, he was often treated as an infallible seer. As one former GE staffer told Fortune writer Geoff Colvin, “When the top guy is the smartest guy in the world, you’ve got a real problem.”⁴ Immelt never claimed to be all-knowing, but bureaucratic power structures invariably cast the CEO as a superhero—a myth that’s often willingly perpetuated by reverential employees, star-struck journalists, and fawning consultants.

The point is, assumptions of exaggerated executive competence are endemic to bureaucracy—a fact that undermines the quality of decisions and, over time, erodes the confidence of employees in their leaders.

Misjudged Competence

However much we may struggle to be objective about our own capabilities, we score even worse when it comes to judging the abilities of others. Research shows our assessments usually say more about us than those we’re evaluating. Again, this phenomenon has its own name—idiosyncratic rater bias. Three factors, in particular, sabotage our ability to reliably assess others.

First, some of us grade tough, while others are consistently generous. In three studies conducted between 1998 and 2010, managers, peers, and subordinates were asked to rank the performance of their colleagues. On average, more than 60 percent of the variation in ratings could be traced to the rating style of the evaluators.⁵ These differences make individual assessments highly unreliable.

Another distortion comes from the fact that we tend to rate most highly those who are most like us. Much as we might wish it otherwise, we tend to divide the world into “us” and “them”—native-born versus immigrant, conservative versus liberal, believer versus nonbeliever, and beautiful versus plain. Psychologists call this “in-group bias.” Despite our enthusiasm for diversity, in-group biases are deeply rooted and are observed even in preverbal children. In one study, eleven-month-old babies were given the chance to choose between two snacks, graham crackers or Cheerios, and then offered two puppets, one of which expressed a preference for the child’s favorite snack, and one which chose the alternative. By a four-to-one margin, babies opted to play with the puppet that shared their culinary preference.⁶

As woke adults, we’re more conscious of our biases, but it’s still hard to disentangle the question of “who’s competent?” from the question of “who makes me feel comfortable?” In her book, Brotopia, Emily Chang notes that while Wall Street banks employ roughly the same number of men and women, women hold only 25 percent of tech industry jobs.⁷ Worse, women attract a minuscule 2 percent of venture funding. While most tech leaders claim to be all in on meritocracy, the evidence suggests that excellence counts most for those who’ve already passed the “bro-hood” test. This sort of insidious in-group bias produces what software pioneer Mitch Kapor calls a “mirror-tocracy.”⁸

There’s another cognitive quirk that leads to misjudgments—the halo or horns effect. As human beings, we’re prone to judge others hastily, often on the basis of first impressions. These initial opinions are resistant to change, even in the face of new data. Researcher David Schoorman found that the biggest factor impacting an employee’s performance review was whether or not he or she had been hired by the person doing the evaluation.⁹ Thanks to the halo bias, a favored deputy may underperform for months or years before getting the boot.

The corrosive effects of these biases are exaggerated by the fact that judgments about an individual’s competence are often dependent on the views of a single assessor—the employee’s boss. In a poll conducted by consultant John Gardner, more than three hundred executives were asked about the prevalence of favoritism in promotional decisions.¹⁰ For the purposes of the study, favoritism was defined as “preferential treatment based on factors unrelated to a person’s abilities, such as background, ideology or gut instincts.” Gardner’s study revealed:

• Seventy-five percent of the executives had witnessed favoritism in hiring decisions
• Ninety-four percent believed policies aimed at preventing favoritism were ineffective
• Eighty-three percent said favoritism produced poor-quality promotion decisions

Put simply, the “data” used in hiring and promotion decisions is riddled with bias—and everyone knows it. In a study conducted by the Corporate Executive Board, 77 percent of HR executives conceded that typical assessment methods don’t accurately measure employee capabilities and contributions. A separate CEB study found zero correlation between individual performance ratings and actual business results.¹¹ That’s about as uncorrelated as you can get.

While many HR professionals recognize the need to overhaul performance management, the usual fixes—abandoning forced rankings, moving the process online, and creating more frequent opportunities for assessment—do little to counter systematic bias.

Overweighted Competence

Among the panoply of skills that are critical to an organization’s success, bureaucracy elevates one above all others: administrative expertise. What distinguishes managers from nonmanagers is not creativity, foresight, or technical expertise, but their the mastery of administrative arcana—developing plans, building budgets, doling out tasks, preparing reports.

Admittedly, there are a certain number of administrative tasks that have to be performed in any organization, but as rule, this work is tangential to the creation of competitive advantage. It’s not administrative competence that generates a patent, spawns a new product, or reimagines a business model. We’re not saying that managerial work is unimportant; it’s vitally important, and when done badly can bring a firm to its knees. As a rule, though, administrative competence is unlikely to lift a company above its peers. It is to organizations what breathing, eating, and sleeping are to human beings—necessary, but not pivotal.

There was a time, several generations past, when administrative skill was rare, but as we’ll argue in chapter 16, this is no longer the case. Nevertheless, in the United States, managers and administrators take home 30 percent of all wages and salaries, despite making up only 18 percent of the workforce.

In a bureaucracy, compensation correlates with rank. In a Fortune 500 company, an executive vice president may make $5 million a year, while a vice president, two rungs down earns a comparatively measly $500,000. In theory, this multiple reflects differences in the difficulty and impact of the work performed. In practice, such differences are often more imagined than real. While an executive vice president (EVP) is likely to oversee a bigger organization than a VP two levels down, this alone doesn’t make the EVP’s job more difficult. To take a hypothetical case, it’s not obvious that the work of overseeing a thousand employees spread across dozens of regional sales teams is intellectually more taxing than leading a one-hundred-person product development team. As a rule, EVPs aren’t solving partial differential equations while their subordinates are struggling with long division—and yet, they’re often paid as if they are.

You might argue that the decisions of an EVP are likely to be more momentous than those of a lowly VP, but even if that’s the case, a yawning salary differential would be justified only if the senior executive was demonstrably more sagacious than his or her subordinates. Unfortunately, there’s little evidence that wisdom correlates with rank. Indeed, a growing body of research suggests the opposite—that positional power increases the odds of bone-headed decisions. Dacher Keltner, professor of psychology at UC Berkeley, has spent more than two decades studying the effects of power. His conclusion: “Power makes individuals more impulsive [and] less risk-aware.”¹² In other words, while an EVP’s decisions may be more consequential than those made by lower-level managers, they’re no more likely to be right, and when they’re wrong, they’re really wrong. That’s why we argued in the previous chapter that, whenever possible, big decisions should be vetted by the crowd.

In short, administrators enjoy a disproportionate share of power and financial emoluments not because their work creates a disproportionate amount of value, is more challenging, or is more likely to be “on the money,” but because bureaucracies tend to overvalue administrative competence, and to pay managers based on the size of their budget or headcount rather than on their net value-added.

Again, it doesn’t have to be this way. Vanguard organizations like Haier, Nucor, Vinci, W.L. Gore, and others distribute a substantial share of administrative work to frontline employees. You’ll recall that Haier shed ten thousand middle management jobs when it moved to its microenterprise model—a shift that enhanced rather than impaired its organizational effectiveness.

Earlier we mentioned GE’s jet engine plant in Durham, North Carolina, noting that the facility employed three hundred technicians and only a single senior administrator—the plant manager. In the hour-long overlap between shifts, teams huddle in conference rooms to review production plans, resolve supply chain issues, adjust work assignments, review productivity data, and work through HR issues—all without the oversight of any formally titled managers.

Hard as it may be to admit it, in a meritocracy, management is one skill among many, rather than one skill to rule them all.

Toxic Competence

In chapter 3 we described bureaucracy as a massive multiplayer game in which employees compete for the prize of promotion. In these tournaments, there’s a single winner—a lone contestant who gets bumped up to become a manager, department head, or VP. Ideally, promotion testifies to an individual’s superior leadership skills or technical knowledge. In practice, promotion often rewards those who’ve mastered the dark arts of bureaucratic combat: hoarding talent, ducking tough decisions, deflecting blame, undermining rivals, and brownnosing the boss.

In a bureaucracy, megawatts of emotional energy get wasted on petty battles, data gets weaponized against adversaries, collegiality gets shredded by zero-sum promotion tournaments, and decisions get corrupted by artfully concealed self-interest. As we’ve noted before, and will again, bureaucracy doesn’t bring out the best in people, nor does it reliably get the best people to the top.

To change all this, to replace bureaucracy with meritocracy, we must do four things: decontaminate judgments about merit, better align wisdom and authority, match compensation to contribution, and build natural, dynamic hierarchies. Let’s take each in turn.

Decontaminating Judgments about Merit

Despite its struggles to hire more women and minorities, Google has long been committed to the idea of meritocracy. The company hasn’t eliminated traditional reporting structures, but it does take pains to reduce managerial bias. This starts with the hiring process. Outside candidates for the position of team leader or above get interviewed by at least four individuals: the manager seeking to fill the slot, a peer of the hiring manager, a representative from a different department, and one or two direct reports. Each interview contributes equally to a candidate’s rating. Those who pass these in-person interviews are further vetted by hiring groups at the departmental and senior leadership level.

Promotions are made by cross-unit groups that rely heavily on feedback from peers and subordinates. In a bid to ensure objectivity, every candidate’s qualifications are benchmarked against the profiles of those who’ve recently been promoted into similar roles across the company.

Performance reviews are similarly broad-based. Every year, colleagues rate one another’s work in an online survey. Subsequently, groups of five to ten senior leaders meet to compare the distribution of ratings within and across teams. This process reduces the pressure managers might otherwise feel to inflate their team’s scores and reveals idiosyncracies in how teams are rated.

By reducing the influence of individual managers in hiring, promotion, and performance reviews, Google minimizes bias and favoritism while making it clear that competence counts for more than gamesmanship. Laszlo Block, Google’s former head of people operations, argues that this approach “sends a strong signal to candidates about Google being nonhierarchical, and it also helps prevent cronyism.”¹³ As a Googler, you know your career isn’t in the hands of your boss. Instead of wasting time sucking up, you can focus on doing great work.

Connecticut-based Bridgewater Associates, the world’s largest hedge fund, has taken an even more radical approach to building a meritocratic organization. With $160 billion under management, the company’s fifteen hundred team members are charged with producing superior returns by making bets on macrotrends like inflation, exchange rates, and GDP growth. Bridgewater’s flagship fund, Pure Alpha, generated $45 billion in investor returns between 1991 and 2015—an industry record.¹⁴

Ray Dalio, the son of a jazz musician, started Bridgewater in his two-bedroom New York City apartment in 1975. In his book, Principles, Dalio writes that the company operates as “an idea meritocracy, not an autocracy in which I lead and others follow, and not a democracy in which everyone’s vote is equal, but a meritocracy that encourages thoughtful disagreements and explores and weighs people’s opinions in proportion to their merits.”¹⁵

To operationalize the notion of a meritocracy, Bridgewater developed the “Dot Collector”—a real-time feedback app that gives employees the opportunity to rate one another on a one-to-ten scale across more than a hundred attributes, such as “learns from mistakes,” “diagnoses root causes,” “thinks strategically,” “demonstrates intellectual horsepower,” “exercises creativity,” “is a meticulous problem solver,” and “proactively shapes change.”

Team members are encouraged to use the Dot app throughout the day as they interact with one another. A twenty-four-year-old junior associate participating in an investment meeting with Dalio is expected to be as honest in evaluating the company’s founder as senior leaders. (Twenty percent of the dots Dalio receives are ratings of four or below, which is considered negative feedback.) Over the course of a year, a typical associate will garner more than two thousand dots—or roughly eight per day.¹⁶ Senior leaders often rack up many times that amount.

Open the Dot app, and you’ll see your average rating across ten broad areas, such as “practical thinking,” “management skills,” and “determination.” Double-click on a category, and the app reveals the ratings you’ve received within the subcategories. Each rating shows up as a color-coded dot along a timeline. (Green dots correspond to ratings of seven or above and red dots to ratings of five and below.) Click on a dot, and you can see who rated you, and when. You can check out everyone else’s ratings as well.

Not surprisingly, Dot profiles get intensely scrutinized when making staffing decisions. A typical case involved the question of whether or not to promote an interim department head to a full-time role. While the candidate was convinced he had the right abilities, others were less sure. Instead of the CEO adjudicating the matter, the interested parties gathered in a conference room and threw the candidate’s Dot scores up on a screen. Dalio recounts the experience: “We stared hard at it together. We then asked the employee to look at that body of evidence and reflect on what he would do if he were in the position of deciding whether he’d hire himself for the job. Once he was able to step back and look at the objective evidence, he agreed to move on and try another role at Bridgewater more suited to his strengths.”¹⁷

Maybe the idea of an “always-on,” hypertransparent review process makes you queasy, but the Dot Collector isn’t as radical, or unique, as it seems. Most university professors get reviewed by their students at the end of every semester. Detailed feedback is collected online and can be easily viewed by other students and faculty. Though disconcerting to some, this sort of open, peer-based review is a far better barometer of competence than a once-a-year, top-down performance review. Bridgewater’s approach highlights expertise, improves the fit between aptitude and responsibility, encourages leaders to be honest about their limits, and creates incentives for personal growth. Most of all, it reduces the risk of single-rater bias. That makes the Dot Collector an essential tool in creating an honest appraisal of individual capabilities.

Aligning Wisdom and Authority

In a perfect world, influence would correlate with expertise rather than positional power, and would be contingent on the topic at hand. Here again, a process like Bridgewater’s Dot Collector pays dividends. Transparent and nuanced competence data is a powerful tool for determining how to weight competing views on a particular decision.

Consider the debate within Bridgewater’s investment team at the height of the European debt crisis in 2012. There were some who expected the European Central Bank (ECB) to break with precedent and buy large chunks of sovereign debt from countries like Italy, Ireland, and Spain. Others thought the ECB would line up behind Germany, which was opposed to a bailout. Hours of debate surfaced compelling arguments on both sides, and a poll indicated a virtual deadlock. As a final step, the opinion of each team member was assigned a credibility score based on their relevant Dot Collector ratings. It quickly became clear that those who were the most credible thought the ECB would print money to buy government debt. That judgment became the investment team’s consensus and was proven right a few days later when Mario Draghi, ECB president, announced the bank would do “whatever it takes” to save the euro.¹⁸

This is how most decisions are now made at Bridgewater, where influence is a product not of tenure or title but of an individual’s peer-attested “believability.” In Dalio’s view, believability-based decision making …

Eliminates what I believe to be one of the greatest tragedies of mankind, and that is people arrogantly, naively holding opinions in their minds that are wrong, and acting on them, and not putting them out there to stress test them. Collective decision-making is so much better than individual decision-making if it’s done well. It’s been the secret sauce behind our success. It’s why we’ve made more money for our clients than any other hedge fund in existence and made money 23 out of the last 26 years.¹⁹

Dalio claims that in his forty-five years at Bridgewater, he’s never made a decision contrary to the believability-weighted advice of his peers, because “to do so [would be] arrogant and counter to the spirit of the idea meritocracy.” For Dalio, the risk of reverting back to positional authority is that he’d “lose both the best thinking and the best thinkers, and … be stuck with either kiss-asses or subversives who kept their disagreements and hidden resentments to themselves.”²⁰

It’s hard to argue with Dalio’s point that “power should lie in the reasoning, not the position, of the individual.” Whatever the approach, there’s a pressing need for decision processes that better align expertise and authority.

Matching Compensation to Contribution

If wisdom doesn’t correlate with rank, neither should compensation. Google gets this. The range of rewards for Googlers working at the same level often varies by more than 300 percent.²¹ A few particularly capable engineers are rumored to have multimillion-dollar pay packages, based on their ability to improve the speed and efficacy of Google’s algorithms.²² As Google’s then chairman Eric Schmidt wrote with Jonathan Rosenberg in How Google Works: “What’s most important in the Internet Century is product excellence, so it follows that big rewards should be given to people who are close to great products and innovations. Pay outrageously good people outrageously well, regardless of their title or tenure.”²³

Compensation at W.L. Gore, the maker of Gore-Tex and more than a thousand other high-tech products, is similarly divorced from rank. Once a year, every associate is asked to compile a list of five to twenty colleagues who have firsthand knowledge of their work. These nominations are then used in a peer-rating process based on pairwise comparisons. By way of example, assume that Tom and Rebecca both list Jennifer as a potential reviewer. In this case, an algorithm will identify the match, and Jennifer will be asked to indicate which of her associates, Tom or Rebecca, contributed more to Gore’s success over the preceding year. (Contribution is defined as the extent and nature of one’s impact on business results.) Tens of thousands of such comparisons are collected across the company and aggregated to create a contribution ranking for every associate. Once the ratings are in, local contribution committees review the results and, when appropriate, fine-tune the rankings. For example, if an associate received substantially higher rankings from top performers than her overall ranking indicates, her position might be nudged upward. Each local committee includes an “equity champion” who’s responsible for alerting the committee to potential biases.

Armed with the rankings, the committees then review compensation data. The goal is to ensure that an individual’s pay reflects his or her peer-derived rating and stays in sync with the pay of similarly rated peers. If the average pay raise in a given year is 4 percent, a highly ranked associate might get a 15 percent increase, while a poorly ranked associate would get no raise at all. Global and regional compensation committees focus on specific functions such as engineering, production, and finance, and review the results to ensure they’re appropriately calibrated across the enterprise and with external benchmarks.

Gore’s peer-based compensation system pushes everyone to think about how they could add more value. The system also encourages collaboration. At Gore, associates understand they report to their peers, not a boss, and are thus more inclined to go the extra mile for colleagues.

Though their approaches are dramatically different, both Google and Gore work hard to ensure that compensation reflects contribution, not rank. They want the energies of every employee to be invested in building a better business rather than winning a promotion tournament.

Building Natural, Dynamic Hierarchies

The idea of meritocracy doesn’t negate the value of hierarchy. As noted earlier, depending on the topic, some individuals deserve to have more authority than others. Not everyone is equally competent and/or believable. The problem with bureaucracy isn’t hierarchy per se, but the dominance of a single, formal hierarchy. In the traditional pyramid, power is vested in positions—it’s binary and allocated top-down. This creates perilous pathologies.

Getting Started

Whether Morning Star, Gore, Haier, or Bridgewater, the point is the same: you can’t build a robust meritocracy until the formal hierarchy gives way to natural hierarchies that are less imperious and rigid.

Here’s a short menu for building a genuine meritocracy in your organization:

1. As a start, ask your peers to rate your expertise across a range of categories, as well as your value added. Share your ratings with those in your network and ask them for advice on how you can improve. Invite others to follow your lead.
2. More generally, ensure that competence and performance ratings are peer-based, with at least five assessors for every individual. Make these ratings transparent to all.
3. Give significant weight to peer assessments in all hiring and promotion decisions.
4. Wherever possible, divorce compensation from rank and tie it more closely to peer-based ratings.
5. Redesign decision processes to give a greater share of voice to those with relevant, peer-attested competence. Downgrade the influence of positional power in decision making.
6. Give teams the right to “fire” incompetent or tyrannical leaders.
7. Finally, create more opportunities for individuals to become meritorious. Rotate team members across roles, challenge people with stretching assignments, open up management training to frontline team members, and take time to mentor others.

The goal of humanocracy is to create an environment in which everyone is inspired to give their best. That won’t happen as long as a significant share of individuals in an organization believe that it’s the blowhards who get ahead, that their own capabilities and contributions are often misjudged, that the suits get an excessive share of the spoils, and that many of their leaders aren’t actually worth following. The antidote to these poisonous realities is meritocracy—a principle that is central to the work of creating human-centric organizations.