A Debate about the Promotion Paradox

[00:00:00] Welcome to the debate. What if I told you that the absolute worst thing you could do for your company's top salesperson is to, well, promote them to run the sales team? Oh, it's a classic trap. Right? I mean, if anyone listening has ever watched their brilliant, like, hyper-competent tech lead or top billing account executive suddenly become this deeply anxious, micromanaging, and ultimately destructive boss, you have witnessed exactly what we are unpacking today. We're pulling apart the promotion paradox.

[00:00:27] Yeah. And this is that deeply entrenched habit that organizations have of promoting individuals into management roles for reasons that, frankly, completely fail to predict their actual leadership capability. Totally fail. Right. So we are drawing today on Dr. Jonathan Westover's recent synthesis of the data. He looked at decades of economic research and organizational psychology, and he found this massive persistent gap between who looks like a leader and who actually leads effectively.

[00:00:54] I mean, the financial and human costs of this paradox are just staggering. They really are. And, you know, we are going to explore the mechanics behind this failure just to establish where we stand on this issue. I argue that this paradox is fundamentally a crisis of measurement and evaluation. Like, organizations promote based on flawed, highly visible signals, and this must be corrected through rigorous, structured, blind assessment. And I take the position that this is a crisis of organizational architecture and incentives.

[00:01:20] Basically, organizations promote the wrong people because management is structurally the only mechanism available for rewarding top performers. Well, let's start by looking at the mechanics of this measurement crisis, because the root of the promotion paradox really lies in a persistent, almost deliberate conflation of two entirely different concepts. Right? Leader emergence and leader effectiveness. Two very different things. Yeah. Exactly. Leader emergence is simply about who looks the part.

[00:01:47] It's, you know, the person who speaks up first in meetings, the person who projects total certainty. The one who's always raising their hand. Yes. The one who physically raises their hand and volunteers to take charge. But leader effectiveness, on the other hand, is about whether the team actually performs under their guidance. And organizations currently rely on highly unstructured judgment. Because executives are time poor, they rely on heuristics, these mental shortcuts. They suffer from visibility bias, meaning they just reward the loudest voice in the room rather than the most capable mind.

[00:02:16] Yeah, that visibility bias is incredibly pervasive. But I think the distinction between emergence and effectiveness really requires a hard look at the data to fully appreciate. Oh, absolutely. And the 2026 study in the Quarterly Journal of Economics by Weidman and colleagues isolates just how dangerous this conflation actually is. They ran this massive controlled study comparing teams with self-nominated volunteer managers against teams with randomly assigned leaders. Like literally drawn out of a hat? Literally. Managers chosen by a lottery.

[00:02:46] And the self-nominated managers actually led their teams to worse performance than the lottery winners. That is wild. Right? The psychological mechanism here is crucial, though. The traits that make someone eager to lead, so overconfidence and a desire for status, are often the exact opposite of the traits required to build a functioning system and elevate others. But you know, it goes deeper than misplaced confidence. This subjective measurement system is fundamentally biased. How so?

[00:03:13] Well, look at the 2026 American Economic Review study by Benson Lionshu. They looked at nearly 30,000 management track employees at a huge retailer. They found that subjective, quote-unquote, potential ratings accounted for roughly half of the entire gender promotion gap at the company. Wow. Half. Half. And the tragedy of this data is that the women actually had higher actual performance ratings. And they later outperformed the men who had been given the exact same highly subjective potential score.

[00:03:42] That's incredibly frustrating. It is. Our measurement tools, these vague gut-level judgments of potential, they're broken. They are filling our leadership pipelines with individuals who are just good at signaling competence rather than actually possessing it. I mean, that's a compelling argument. But have you considered that subjective measurement is merely a symptom, not the core disease? A symptom? Yeah. I argue that the disease is what we call the performance-to-promotion reflex.

[00:04:10] We cannot look at this strictly as an evaluation error when the actual architecture of the modern corporation makes this error totally inevitable. You know, the Peter Principle. Right. That people are promoted to their level of incompetence. Exactly. It's not just a joke. It is an empirically verified reality. Let's look at a different Benson, Lee, and Shu study. This one from 2019. They analyzed data on nearly 40,000 sales workers across 131 firms.

[00:04:37] And they proved that firms heavily favor their absolute best individual sellers for promotion. Which seems logical on the surface. On the surface, sure. But here is the devastating mechanism. Doubling a new manager's pre-promotion sales is associated with a 6% decline in their subordinate sales. A decline. Yes. Think about what that actually looks like on the sales floor. We reward the person who is best at doing the job by giving them a job where they actively prevent everyone else from doing it. They step in.

[00:05:06] They micromanage. They take over pitches. Because their skills are rooted in individual execution, not system building or coaching. I see why you think that. But let me give you a different perspective. What you just described, the failure to recognize that individual execution and team coaching are completely different skill sets. That perfectly illustrates my point. How so? We are measuring the wrong variables. We are essentially looking at a thermometer to tell us the weight of an object. It's just bad measurement. But why are they measuring the wrong variables?

[00:05:35] It's not because executives are simply ignorant of the difference. It's because they are caught in a structural trap. We really have to talk about the political economy of the corporation. Okay. Explain that. So, the vice president of sales has a quota to hit, right? Their top closer demands a 20% raise and more prestige or else they are leaving for a competitor. Human Resources says the only way to authorize that level of compensation and status is to give them a director title. The organization essentially bribes its best individual contributor with a management role. Right, but...

[00:06:05] Wait, just think about it. You could have the most perfect scientifically validated measurement system in the world. But if the architecture dictates that the only pressure valve for top performer salary demands is the management track, the pipes are going to burst. Until you fix the incentive architecture, you are going to force top performers in the roles they aren't suited for simply to pay them what the market says they are worth. I firmly believe that if we build structured blind assessments, we can strip that visibility bias out of the system entirely.

[00:06:31] We can stop that reflex dead in its tracks, regardless of the internal political pressure. You know, we can take the decision completely out of the hands of the desperate VP of sales. That's a huge if, though. Well, the U.S. Army actually provides a master class in this with their Battalion Commander Assessment Program, or BCAP. For decades, they used a legacy file-based system. They looked at past performance and, you know, subjective evaluations from superiors. Which is exactly what corporations do. Right.

[00:06:55] And that inherently carries the halo effect, where a person's success in a past tactical role just blinds the evaluator to their unsuitability for a future strategic role. But in 2020, the Army shifted to a rigorous five-day assessment. They implemented cognitive and non-cognitive tests, psychological evaluations, and blind panel interviews where the candidates' files were completely anonymized. Completely anonymized? Yes. The panel couldn't see the candidate, couldn't see their past accommodations, and couldn't be influenced by their physical bearing or charisma.

[00:07:23] They also incorporated blind peer and subordinate assessments. I mean, it was a fascinating intervention, fundamentally shifting how they weigh past loyalty versus future capability. Exactly. And the results proved that measurement is the primary lever for change. The Army achieved a 34% change in the officers selected for command compared to their legacy system. That is a massive shift. 34%.

[00:07:47] One in eight candidates who would have historically been handed a battalion based on their past performance were explicitly found not ready when measured objectively. When you force the system to measure the actual psychological and cognitive requirements of the job, without that halo effect of past performance, you completely change who ends up in the seat. You prevent the structural failure before it happens. I come at it from a different way. That Army example is brilliant, really, but it actually proves my point about structural architecture. How does it prove your point?

[00:08:15] Because the military can implement BCAP because rank, role, and compensation in the Armed Forces are rigidly separated in ways that the corporate sector simply hasn't matched. I mean, a major is a major, regardless of whether they are commanding a battalion or serving in a staff role. Sure. The compensation structure is different. It's entirely different. In the corporate world, without genuine architectural parity, your structured assessments are just going to face insurmountable pressure. The corporate machine will absolutely override the assessment to save the star player.

[00:08:43] I'm not convinced by that line of reasoning because, well, if we build alternative career tracks but keep our biased assessment systems intact, aren't we just building two broken escalators instead of one? Two broken escalators? Yes! Because if the evaluation criteria are still driven by who speaks the loudest and who has the most visibility, you'd just be sorting people into two different paths based on the same fundamentally flawed metrics. The bad data is the root problem here. I'm sorry, but I just don't buy that. Let me tell you why.

[00:09:09] When you build the dual tracks, you are removing the artificial gravity that pulls everyone toward the management escalator in the first place. The artificial gravity? Exactly. You have to look at what companies like GitLab and Procter & Gamble have done with dual career ladders. GitLab openly publishes its engineering framework. If you reach the senior level, the track splits. You can choose to become an engineering manager or you can become a staff-level individual contributor. And here is the crucial architectural intervention.

[00:09:37] A staff engineer has the exact same base compensation, the exact same equity potential, and the exact same prestige as the engineering manager. P&G pioneered this decades ago in their R&D departments. When you structurally decouple the reward from the role, when your top seller or top coder can make manager-level money without having to manage a single human being. You remove the incentive to jump tracks just for the cash. You instantly remove the desperate panic drive for the management title.

[00:10:03] People stop self-nominating for roles they don't actually want just to get a raise. You remove the pressure valve entirely. But that still assumes that once they choose the management track, the organization will naturally select the right people without rigorous intervention. Let's shift to what happens when we lack that structured upfront assessment. Because the damage isn't just a slight dip in productivity. We really need to talk about the consequence gap. The consequence gap, right. Yeah.

[00:10:28] When unstructured judgment allows the wrong person, particularly a person with toxic traits, to gain institutional power, the damage is catastrophic. The 2026 G-Check report surveyed thousands of workers about toxic colleagues. Now, when the toxic colleague was a junior employee, 41% of respondents said the person was removed or left, and only 24% said they were protected. Which makes sense. Right. But when that toxic colleague was a manager or senior leader, the script flipped entirely.

[00:10:55] 36% of respondents said the toxic manager was actively protected or rewarded, and only 21% were removed. Wow. The mechanism here is institutional armor. When you allow people to achieve a management position based on raw sales numbers or charismatic signaling, the organization will protect them to protect its own revenue stream. This proves that an upfront objective structured filter is absolutely critical. We have to filter out these destructive traits before they acquire the institutional power that makes them untouchable.

[00:11:23] See, you are looking at the consequence gap and concluding that the upfront filter is the missing piece. I look at the consequence gap and see a catastrophic failure of ongoing upward accountability. Upward accountability. Yes. The problem isn't just who gets through the door. The problem is that once they are in the room, they aren't measured by the people they are actually supposed to be leading. Leaders must be measured by their teams, and those metrics must have actual structural teeth.

[00:11:50] Upfront filters, no matter how rigorous, are just point-in-time predictions. Upward accountability is continuous reality. But do companies actually enforce that? They do if it's structural. Look at Google's Project Oxygen. They didn't just guess what made a good manager. They analyzed their own data. They found that technical expertise, the very thing that usually gets people promoted, ranked dead last among the behaviors of their best managers. Dead last? Dead last.

[00:12:17] The top behaviors were coaching, empowering the team, and not micromanaging. Sure, but knowing those behaviors doesn't naturally translate to enforcing them across a massive corporation. It does if it's integrated into the architecture. Google didn't just screen for coaching up front. They built an upward feedback survey directly measuring those specific behaviors, and they tied it to the manager's performance review. Between 2010 and 2012, median favorability scores for managers rose from 83% to 88%, and the lowest scoring managers improved the most.

[00:12:47] The mechanism here is consequence. Okay, that's tech, though. Well, look at the medical field, then. The Mayo Clinic surveyed physicians about their supervisor's leadership behaviors. They found that a single-point increase in a supervisor's leadership score on a 60-point scale was associated with a 3.3% decrease in the likelihood of physician burnout and a 9% increase in satisfaction. Now, that's a significant shift from just one point. Exactly. Yeah. And we aren't just talking about a slightly happier workplace here.

[00:13:16] When we talk about physician burnout, we are talking about medical errors, patient safety, and lives on the line. The organization must structurally hold managers accountable to the retention and well-being of their team as a primary metric of success. If you build that architecture, the consequence gap collapses because the toxic manager's behavior instantly impacts their own performance metrics. Okay. And, you know, this brings me to a critical shift in the modern workplace.

[00:13:40] The 2026 Gartner survey of nearly 3,000 employees and managers shows us that the role of the manager has fundamentally transformed. 47% of managers say more is expected of them now than just a year ago. Oh, definitely. The job is much harder now. Right. The survey found they are spending roughly nine hours a week just dealing with employees' personal, emotional, and well-being concerns. The mechanism of management has shifted from simply assigning tasks and monitoring output to providing deep psychological scaffolding.

[00:14:09] You could not simply use a better sorting hat at the beginning of the process to solve for this. A sorting hat? Yeah. You have to structurally support them. Look at the Wallace Foundation's principal pipelines in education or Walmart's manager academy. Walmart takes their newly promoted store managers, sends them to a week of intensive in-person training at headquarters, and pairs it with a six-month structural mentorship. And it works? Yes.

[00:14:33] The Wallace Foundation showed that schools with newly placed principals from their supported pipelines outperform comparison schools significantly in math and reading. You have to build an architecture of preparation and ongoing support because the role is now fundamentally one of coaching. That's an interesting point, though I would frame it differently. What you're describing, the increasing need to navigate emotional complexities, mitigate burnout, and handle ambiguous human crises, that actually makes the case for rigorous upfront assessment even stronger.

[00:15:03] Stronger? How? Because if managers are spending nine hours a week on deep emotional and cognitive support, relying on past individual performance or fluid self-nomination is even more disastrous than it was a decade ago. I don't disagree that support is necessary, but I fundamentally reject the idea that we can train our way out of a bad selection. I'm not saying we train bad selections. I'm saying... Let me bring us back to the Weidman study from 2026. They didn't just find that self-nominated managers failed.

[00:15:30] They identified the actual empirical predictors of team leadership. And it wasn't competent. It wasn't past sales. It wasn't demographic characteristics. The true predictors were measured skill in economic decision-making and fluid intelligence. Fluid intelligence being the capacity to adapt rather than rely on crystallized experience. Exactly that.

[00:15:51] Fluid intelligence is the ability to connect the dots in completely novel situations, to synthesize conflicting data streams, and to solve problems you've never seen before without a playbook. It is cognitive agility. When a manager is dealing with a team in crisis, they cannot rely on the crystallized knowledge of how they closed a software deal five years ago. They need fluid intelligence. Right, but... These are baseline cognitive and behavioral traits. You cannot send someone to a six-month mentorship and teach them fluid intelligence.

[00:16:20] We absolutely must use structured assessments, like the Army's BCAP, to test for these specific capabilities before placing anyone in such a high-stake environment. If you put someone in a role who lacks the baseline cognitive capacity for complex human leadership, all the architectural scaffolding in the world will just be propping up a collapsing building. But if we acknowledge that fluid intelligence and agile decision-making of the true predictors, we are right back to the structural problem. Why?

[00:16:45] How do we get the organization to care about those baseline metrics when the internal political economy is screaming at them to retain the top closer? You keep focusing on the purity of the evaluation, but you are totally ignoring the reality of the corporation. The organization will quietly bypass your brilliant assessment tools if they don't have another mechanism to compensate that top performer. Well... The structural decoupling of compensation and prestige from the management track is the absolute prerequisite for your assessments to even be allowed to function.

[00:17:13] The assessment functions precisely because it removes the discretion of the biased actors. Only if they let it. When you institute something like Amazon's bar raiser program, where highly trained, independent assessors who are not part of the hiring team hold an absolute veto over the promotion decision, you break the internal political economy. You take the decision away from the desperate sponsor.

[00:17:34] In the Hoffman, Khan, and Lai study from 2018, they analyzed what happens when managers override structured job tests using their gut instinct or private information. And? The data showed that when human discretion overrides the assessment, they consistently make worse hires, resulting in higher turnover and lower productivity. Human discretion in these promotions is inherently flawed. If the organization mandates the assessment and mandates strict adherence to the outcome, the internal politics are forced to adapt to the new reality.

[00:18:04] But mandating adherence by using an independent veto is itself an architectural intervention. You are proving my point that structure dictates behavior. No, I'm saying the measurement has to be the gatekeeper. But let's step back and look at what we actually agree on here. Because our convergence is just as illuminating as our divergence. We are in absolute agreement that the traditional corporate model, promoting the best individual contributor simply as a prize for past performance,

[00:18:31] or rewarding the person who signals the most confidence, is an actively destructive practice. Oh, it is deeply destructive. The empirical evidence is overwhelming that a highly capable manager is a massive multiplier for team output, while an incapable one is an absolute anchor. However, the Lazider, Shaw, and Stanton study, which utilized incredibly granular productivity data,

[00:18:52] showed that replacing a boss in the bottom 10% with one in the top 10% increased a team's total output by more than adding a tenth worker to a nine-person team. That is staggering. It is. One good manager generates more raw output than hiring an entire additional human being. And the impact on retention is just as profound. The Hoffman and Tadalis study showed that moving from a 10th percentile manager in people skills to a 90th percentile manager predicted a 60% reduction in employee turnover.

[00:19:21] I mean, a good manager is the single greatest retention tool a company possesses. And a bad manager is the single greatest threat to its human capital. Where we diverge is really just the sequence and mechanism of the cure. I maintain that rigorous, structured, and debiased assessment is the ultimate key to breaking the promotion paradox. We have the psychological and analytical tools, you know, blind panels, job analysis-based simulations, structured peer reviews,

[00:19:48] to ensure we are measuring true capability, fluid intelligence, and complex decision-making rather than mere confidence and extroversion. If we accurately measure and select the right materials, we will build a stable house that won't collapse under the weight of modern managerial demands. Right. And I maintain that until organizations fundamentally alter their architecture, structurally decoupling compensation, prestige, and influence from the management track,

[00:20:17] even the most scientifically validated assessments will inevitably be bypassed or overridden by the desperate need to retain star players. Organizations must build dual career ladders first. They must integrate upward feedback into performance reviews, ensuring that a manager's success is defined directly by their team's well-being and output. Until you change the structural incentives of the house, it doesn't matter how perfectly you measure the materials. You will still be forcing the wrong pieces into the wrong places.

[00:20:46] Well, the stakes of getting this right are immense, both for the bottom line of the organization and the daily lived psychological experience of the employees. Leadership is not a prize to be handed out. It's not a reward for being a great salesperson, a brilliant engineer, or a loud voice in a meeting. It is a distinct craft with distinct, complex demands. It absolutely is. In the source material we've explored today, Dr. Westover's synthesis of the Weidman, Benson, and Hoffman studies,

[00:21:15] along with the G-check and Gartner data, provides a clear, evidence-based roadmap for professionals and organizations to critically evaluate their own practices. We highly suggest that you explore the source material to evaluate how your own organization handles this paradox. You will have to decide for yourself whether the measurement and assessment protocols must be fixed first, or if the structural architecture and incentive systems require their first overhaul.

[00:21:40] Either way, you cannot build a sustainable bridge by picking the shiniest material and simply hoping it holds the weight. Leadership requires deliberate, intentional design. Thank you for joining us for this intellectual exchange. We will see you next time.