Developing Leaders, Not Followers
Your organization's results are the organization's product; other leaders are yours — and the only honest measure is whether the work is still standing four generations after you.
From the Founder
The hardest thing about developing somebody is that it is inconvenient on purpose. It is faster to do the work yourself. It is faster to make the call yourself. Every hour you spend walking a younger leader through why you decided what you decided produces nothing you can show anybody this quarter. I have learned to treat that as the cost of the job, not a distraction from it. Leadership is doing what is right for the people you lead even when it is not convenient for you, and sometimes that means putting yourself on the back burner — your speed, your credit, your comfort. Here is the part nobody warns you about: the people you develop well will eventually see what you missed and say so. If your identity is settled, that day is the payoff. If it is not, it is a threat, and you will quietly stop developing people right before it arrives.
Executive Summary
Delegation moves work; development changes the person the work moves to. This lesson makes the case that a leader's true product is other leaders, and gives you the Four-Generation Test drawn from 2 Timothy 2:2 — Paul, Timothy, faithful people, others also — as the diagnostic. You will study how Jesus concentrated on twelve while crowds came and went, why the teacher's paradox means success is measured by being surpassed, and what the research on mentoring, feedback, and deliberate practice honestly supports once selection effects and practitioner folklore are stripped out. You will meet Barnabas, who developed the man Paul had written off. And you will confront the executive version: a mayor with no bench is a single point of failure with a good reputation.
Learning Objectives
- Distinguish delegation from development, and explain why an organization can be perfectly delegated and still have exactly one leader
- Apply the Four-Generation Test from 2 Timothy 2:2 to your own leadership, including the generation most leaders cannot name
- State honestly what the research on mentoring, feedback, and deliberate practice supports, and label 70-20-10 as practitioner folklore rather than finding
- Treat succession and bench depth as a first-order executive duty rather than a late-career administrative chore
Teaching Manuscript
The Question 7.4 Left Standing
Lesson 7.4 ended by asking what you are still holding that belongs to somebody else. If you did the work honestly you produced a list — decisions, approvals, relationships, whole categories of judgment that you have been keeping because keeping them felt faster. Good. Now take the question one turn further, and this turn is the one that separates managers from leaders.
Imagine you delegated every item on that list perfectly. Clean handoffs, real ownership, outcomes rather than tasks, no takebacks. Picture the organization eighteen months later. Everything is getting done. Nobody is waiting on you for approvals. And now answer this: how many leaders does that organization have? Because it is entirely possible to run a beautifully delegated operation that contains exactly one leader and a very efficient group of people executing his judgment.
That is the distinction this lesson turns on. Delegation asks who should own this outcome. Development asks who this outcome is going to make. It is the same assignment either way, and the difference lives in what you do after you hand it over. A delegator inspects the result. A developer debriefs the reasoning — what did you consider, what did you rule out, what would have changed your mind, what will you do differently. Same meeting on the calendar, completely different product coming out of it.
The executive stake here is not sentimental. A leader with no developed leaders is a single point of failure with a good reputation. We celebrate those people. They are the ones who can be reached at midnight, who catch the error everybody else missed, whose fingerprints are on every decision that went right. And their organizations are one illness, one scandal, one better offer away from disarray. Understand what I am saying, because it is uncomfortable: being indispensable feels exactly like competence, and it is usually a design failure. Sometimes it is worse than that. Go back to Lesson 1.1 and the fourth counterfeit foundation — the leader whose identity is fused to the position cannot develop a successor, because every capable person he raises is a threat to the only thing holding him up. He will not describe it that way. He will say the bench is not ready. He will have said that for nine years.
So here is the claim, stated plainly. Results are the organization's product. Other leaders are yours. If you are producing results without producing leaders, you are borrowing capacity from the future to make the present look good, and somebody is going to have to repay that loan, probably right after you leave. Stop for a second and answer honestly, on paper if you can: name the people who are leaders today because of deliberate investment you made in them. Not people who worked for you. Not people you liked. People who lead now, in part, because of what you did on purpose. If the list is short, you already know what this lesson is for.
Four Generations in One Sentence
Paul is at the end of his life, writing to a younger man he trained, and he compresses an entire theory of leadership development into one sentence. "The things which you have heard from me in the presence of many witnesses, entrust these to faithful men who will be able to teach others also" (2 Tim. 2:2, NASB 1995). Count the generations. Paul is one. Timothy is two. The faithful people Timothy will entrust it to are three. The others they will teach are four. Four generations, one sentence, written by a man in prison who will not live to meet generations three or four.
The argument rests on the words, so go to them. The verb behind "entrust" is paratithēmi — deposit language. It is the word for placing something valuable into another person's keeping, with an expectation of an account. It is the word on the lips of Jesus at the cross, "Father, into Your hands I commit My spirit" (Luke 23:46, NASB 1995), and the word Peter uses for entrusting one's soul to a faithful Creator. Paul is not describing a lecture series. He is describing a transfer of custody. What you know is not yours; it is on deposit with you, and it is supposed to move.
Then the selection criteria, and notice how few there are. Pistos — faithful, reliable, the kind of person whose word predicts their behavior. And hikanos, the word behind "able" — sufficient, adequate, competent to the task, here specifically competent to teach others. Character and capacity. That is the whole filter. Talent is not on the list. Charisma is not on the list. Current performance is not on the list. Most leaders select for the last one and are then surprised when their high performers cannot develop anybody, which should not surprise anyone, because performance and transferability are different capabilities. One more note for accuracy: the noun Paul uses is anthrōpois, the ordinary Greek word for human beings, not andrasin, the word that specifically means males. NASB 1995 renders it "faithful men" in the older inclusive sense of the English word, and honest translation should say so rather than let a four-hundred-year-old English convention do work the Greek is not doing here.
Now watch the same pattern in the life of Jesus, because he had every incentive to optimize for crowds and did not. Mark records that "He appointed twelve, so that they would be with Him and that He could send them out to preach" (Mark 3:14, NASB 1995). Read the order. Being with precedes being sent. Luke tells us that before he chose them he spent the whole night in prayer (Luke 6:12-13). This is a selection decision treated as a matter of enormous weight, and then followed by three years of proximity — meals, travel, arguments, failures, questions answered badly and asked again. And when the crowd thinned out after a hard teaching, John records that many disciples withdrew and were no longer walking with him, and he turned to the twelve and asked whether they were leaving too (John 6:66-67). He did not chase the crowd. He had already decided what he was building, and it was not an audience.
So name the framework and use it. The Four-Generation Test: can you name generation two, generation three, and generation four? Generation two is easy — the people you are personally developing. Almost every leader can produce that list. Generation three is the people they are developing, and here is where nearly everyone goes quiet. If you cannot name a single person your people are developing, then what you have built is a hub, not a chain, and it terminates the moment you do. Generation four you will probably never meet, and that is the point of the whole exercise. The test of your leadership is not what happens while you are watching.
The Teacher's Paradox
There is a strange thing at the center of this work, and you have to make peace with it before the techniques will do you any good. Every other profession measures success by what it produces. Teaching measures success by its own obsolescence. Jesus states it flatly: "A pupil is not above his teacher; but everyone, after he has been fully trained, will be like his teacher" (Luke 6:40, NASB 1995). The phrase "fully trained" is katērtismenos, a participle from katartizō — a verb used for mending nets, setting something back in order, completing what was unfinished. Training in this vocabulary is repair and completion, not information delivery. And it aims at likeness. The finished pupil resembles the teacher, which means whatever you actually are gets transmitted, not whatever you say.
Then he goes further than any leadership book will follow him. "He who believes in Me, the works that I do, he will do also; and greater works than these he will do" (John 14:12, NASB 1995). Interpreters differ on the scope of "greater" — whether it means greater in kind, in extent, or in the reach made possible after the ascension, and honest teaching should say that the question is open. What is not open is the direction. The teacher announces, as good news, that the students will exceed him. Sit with how rare that is. Most leaders will tell you they want their people to surpass them, and most of those leaders have arranged their organizations so that it cannot happen.
Philosophically, this is the teacher's paradox: the success condition of the role is the redundancy of the person holding it. It is also why instruction alone cannot do the job. Michael Polanyi's argument in The Tacit Dimension is that we can know more than we can tell — that a great deal of real expertise is tacit, held in the hands and the judgment rather than in propositions that can be written down and handed over. That is exactly true of leadership judgment. Which risk is the real one. When a vendor is stalling rather than deliberating. When a room has stopped listening. When a number is technically accurate and substantively false. You cannot lecture those. They transfer by proximity — by a person doing real work near a person who already does it well, with running commentary about why. Jean Lave and Etienne Wenger called the general form of this legitimate peripheral participation: the newcomer does genuine work at the edge of a practice and moves inward. It is the oldest knowledge-transfer technology we have, and it predates every classroom in the world.
Steelman the objection, because it is a real one. Apprenticeship reproduces the master's blind spots with high fidelity. Guilds became conservative for exactly this reason: the transmission mechanism that carries craft also carries prejudice, bad habits, and a settled view of what questions are worth asking. A leader developed entirely by you will inherit your instincts, including the ones that are wrong, and will be less likely than a stranger to notice. That is not a small cost, and the answer is not to abandon apprenticeship but to build a second channel into it — a mentor who disagrees with you, an outside assignment, a rotation into a function where your reputation buys nothing. If the only person shaping your successor is you, you have built a copy, and copies degrade.
So put the paradox to yourself now, in the specific. Who is close enough to you, in actual working proximity, to catch how you think rather than only what you decided? If the honest answer is nobody — if every relationship you have is transactional enough that only conclusions travel — that is the finding, and it is fixable, but not by adding a training program. It is fixed by changing who is in the room when you work.
Without scrolling back: name the four generations in 2 Timothy 2:2, and the two selection criteria the verse actually gives.
What the Research Actually Supports
This academy does not oversell evidence, so let me tell you what the science supports and where it stops, because the leadership-development industry is not going to.
Start with mentoring, since it is the most commonly prescribed intervention. The meta-analytic work — Allen, Eby, Poteet, Lentz and Lima in the Journal of Applied Psychology in 2004 is the standard reference — finds that mentored people report better career outcomes than unmentored people. The effects are real. They are also modest, and they are consistently larger for subjective outcomes like career satisfaction than for objective ones like compensation and promotion. And there is a confound that honest reading cannot ignore: mentoring relationships are not randomly assigned. High-potential people are more likely to be selected for mentoring in the first place, which means some unknown portion of what looks like a mentoring effect is a mentee effect. Later work in the same literature also documents that negative mentoring experiences exist and do damage. Anyone who tells you mentoring is transformative is selling something. What the evidence supports is that it helps, somewhat, on average, and that a bad mentor is worse than none.
Feedback is more interesting, and the classic finding is one every executive should know. Kluger and DeNisi's 1996 meta-analysis in Psychological Bulletin found that feedback interventions improved performance on average — and that in over a third of the cases studied, feedback made performance worse. The pattern they identified is the useful part: feedback that directs attention to the task tends to help, and feedback that directs attention to the self tends to hurt, because the recipient starts defending a self instead of fixing a process. Translate that into your next conversation. "You are not strategic enough" points at the person and will usually damage them. "In that meeting you moved to a solution before the problem had been defined, and here is what it cost us" points at the work and can be acted on tomorrow. The difference is not politeness. It is where the attention lands.
Deliberate practice deserves the same honesty. Ericsson and colleagues established in 1993 that expert performance is built through structured, effortful practice with immediate feedback rather than mere experience, and that framework has held up as a description of how skill is built. But the meta-analysis by Macnamara, Hambrick and Oswald in 2014 found that deliberate practice accounted for a modest share of performance variance overall, and very little of it in professional domains — the domains you and I actually work in. Reps matter. Reps are not the whole story, and a leader who promises that ten thousand hours produces a leader is telling you something the data does not support.
Which brings us to 70-20-10, and I want you to hear this labeled correctly: it is practitioner folklore, not a finding. The research underneath it is real and valuable — McCall, Lombardo and Morrison's work at the Center for Creative Leadership, published as The Lessons of Experience in 1988, documented that when successful executives describe what developed them, they overwhelmingly describe hard assignments rather than courses. That is worth knowing. But even that rests on retrospective self-report, which is memory rather than measurement. The tidy ratio — seventy percent experience, twenty percent relationships, ten percent formal training — was appended later as a heuristic and has never had a serious empirical basis. Use it as a priority ordering if you like. Never quote it as a number, and never let anyone build a budget on it in front of you without saying where it came from. Strip all the overselling away and what survives is unglamorous and durable: real work with real consequences, a person who tells you the truth about it in task terms, and enough time for the lesson to land. That is the whole technology. Everything else is packaging.
The Bench Is a First-Order Duty
Scripture gives us a case study in developing the person somebody else had written off, and it is not a tidy one. Acts 15 records that when Paul proposed revisiting the churches, Barnabas wanted to bring John called Mark, and Paul refused, because Mark had deserted them in Pamphylia. Luke's word for what happened next is paroxysmos — sharp disagreement, the root behind our word paroxysm, the same root used later of Paul's spirit being provoked in Athens. This was not a polite difference. They separated, and Barnabas took Mark and sailed to Cyprus.
Notice that Paul's position was defensible. Mark had quit once on a hard trip, and Paul was making a risk judgment about a mission where the cost of a second desertion would fall on other people. Reasonable leaders make that call. But Barnabas — whose name Luke elsewhere glosses as Son of Encouragement — made a different bet, spent his own partnership with Paul to make it, and then disappears from the narrative. Decades later, Paul writes from prison: "Pick up Mark and bring him with you, for he is useful to me for service" (2 Tim. 4:11, NASB 1995). Scripture does not tell us who was right in the argument, and honest readers differ. What it tells us is that the man written off became useful because somebody was willing to be inconvenienced by him. So ask the question directly: who have you written off, and how confident are you, really, that you were right? That judgment was made with the information you had on the worst day of that person's tenure.
History supplies a version at national scale. Lee Kuan Yew treated leadership renewal as a governing task with a calendar rather than a retirement problem. Through the nineteen-seventies and eighties his party systematically recruited and moved younger leaders into real responsibility, and in November 1990, after thirty-one years as prime minister, he stepped down and handed the office to Goh Chok Tong, who governed until 2004. Say the honest thing alongside it: Lee's record on press freedom, political opposition and civil liberties is genuinely contested, and this academy is not endorsing the system he built. The transferable principle is narrower and still rare — he did the succession work while he was strong, on a schedule, instead of leaving it to a crisis. Most leaders of his generation and stature did not, and their countries paid for it.
Now the executive translation, and this is where it stops being a virtue and becomes a duty. A mayor with no bench is a single point of failure sitting on top of a government of roughly three hundred thousand employees and more than forty agencies. The Charter has a continuity mechanism — on a vacancy the Public Advocate acts as mayor pending a special election — but a legal succession rule is not a bench. It keeps the seat filled. It does not keep the government competent. The real bench is the deputy mayors and the commissioners, and the appointment and supervision of those people is the single highest-leverage decision set a mayor makes.
Directing a commissioner and developing one are different activities, and most executives only ever do the first. Directing produces compliance and a City Hall bottleneck, because every judgment call routes upward to the only person authorized to make one. Developing produces agencies that can act. Concretely, that means debriefing the reasoning behind a commissioner's decision rather than only auditing the result; putting a commissioner in front of the Council or the press during her own agency's hard week instead of absorbing it yourself at the podium; giving real authority early enough that failure is survivable; and building cross-agency assignments so your people learn a government rather than a silo. And be honest about the limits while you do it, because overpromising authority is its own character failure — you can develop a schools chancellor only within the mayoral control the state legislature periodically reauthorizes, and you cannot build an MTA bench at all, because the MTA is a state authority and does not report to you. Put the Four-Generation Test on your actual calendar this quarter, in writing, with names in all four columns and blanks where you cannot fill them. The blanks are the assignment. Then bring it into 7.6, where we ask a harder question about the people you have developed: will any of them tell you the truth?
In one sentence: why is 70-20-10 a useful reminder but an illegitimate statistic?
Through the Six Lenses
Evidence levels labeled per the Truth & Intellectual Integrity standard.
Biblical
2 Tim. 2:2 (NASB 1995) compresses four generations into one sentence; the verb paratithēmi is deposit language — custody transferred with an expected accounting (cf. Luke 23:46). Criteria are pistos (faithful) and hikanos (competent to teach), not talent. Mark 3:14 orders it: with Him, then sent. Acts 15:36-41 records a paroxysmos over John Mark; 2 Tim. 4:11 records the outcome of Barnabas's investment — "he is useful to me for service."
Philosophical
The teacher's paradox: the role's success condition is its own redundancy (Luke 6:40; John 14:12, scope of "greater" disputed). Polanyi's tacit knowledge — we can know more than we can tell — explains why judgment resists instruction and transfers by proximity; Lave and Wenger's legitimate peripheral participation names the mechanism. The strongest objection is real: apprenticeship faithfully reproduces the master's blind spots, which is how guilds calcify. The answer is a second, disagreeing channel, not the abandonment of apprenticeship.
Scientific
Mentoring meta-analysis (Allen, Eby, Poteet, Lentz & Lima, JAP, 2004) finds real but modest benefits, larger for subjective than objective outcomes and confounded by selection — high potentials get mentors. Kluger and DeNisi (Psychological Bulletin, 1996) found feedback improved performance on average yet worsened it in over a third of cases, especially when attention moved to the self. Deliberate practice (Ericsson et al., 1993) is well described; Macnamara et al. (2014) found it explains modest variance and very little in professions. 70-20-10 is practitioner folklore appended to CCL's Lessons of Experience research, not a finding.
Historical
Lee Kuan Yew treated leadership renewal as scheduled governing work, stepping down as prime minister in November 1990 after thirty-one years and handing to Goh Chok Tong, who served until 2004 — the institution outlasted the founder. His record on press freedom and political opposition is genuinely contested and is not endorsed here; the transferable principle is that succession was done from strength rather than in crisis, which most founding-generation leaders never managed.
Influence
Cialdini's commitment-and-consistency research explains why a public developmental commitment — naming someone as an owner in front of peers — changes their subsequent behavior more than private encouragement does. Barnabas illustrates the harder form: sponsorship, where you spend your own credibility on someone unproven, rather than mentorship, where you only spend time. The mentor-versus-sponsor distinction is practitioner literature, not established science, but the cost difference is real and it is the reason sponsorship is rare.
Executive
Drucker treated developing people as a manager's non-delegable duty; Collins's work on Level 5 leadership found that leaders who set successors up to win are the ones whose results persisted. In the mayoral context the bench is deputy mayors and commissioners, and appointing and supervising them is the highest-leverage decision set the office contains. The Charter's succession rule — the Public Advocate acting as mayor pending a special election — fills the seat, not the government. Honest limits apply: schools governance depends on Albany's reauthorization, and the MTA is a state authority.
Case Study
The NYPD, 1994: An Accountability System That Manufactured Commanders
SITUATION. William Bratton became Police Commissioner in January 1994, inheriting a department in which precinct commanders were rewarded for not making waves and had limited authority over their own resources. New York had recorded 2,262 murders in 1990, per the NYPD's published historical citywide crime statistics. CONSTRAINTS. Civil service protections limited wholesale personnel change; the crime problem could not wait for a generation of new hires; and the department's command culture had trained decades of officers to route judgment upward. DECISION. The department built CompStat — recurring sessions where precinct commanders answered personally for their own numbers and their own plan — and paired it with real resource authority at the precinct level. ANALYSIS. The system's second output was leaders. Commanders who could analyze, decide and own a result rose fast, and the department exported chiefs for years afterward — Timoney to Philadelphia and Miami, Norris to Baltimore, McCarthy to Newark and Chicago. Note the honest complication: the same pressure that developed commanders created incentives to shade the reporting, examined in the department's own Crime Reporting Review Committee, which reported in 2013. And the crime decline's causes remain genuinely contested among criminologists; do not adopt a camp. DISCUSSION. In the system you run, who is being developed by the accountability, and who is only being measured by it?
Reflection Questions
- Name generation three: who are the people that the people you are developing are themselves developing? If you cannot name one, what does that tell you about what you have actually built?
- Think of the last person you wrote off professionally. What information did you have, and on what day of their tenure did you have it? Would Barnabas have made the same call?
- When someone you developed disagrees with you publicly and turns out to be right, what happens in your body? Answer honestly — that reaction is the ceiling on how far you can develop anyone.
- If you were removed from your role without notice this month, name the specific decisions that would stall. Each one is an undeveloped person, not a busy schedule.
Practical Exercise — The Four-Generation Map
On one page, draw four columns. Column one: you. Column two: every person you are deliberately developing right now — name them, and beside each name write the specific assignment carrying the development. Column three: for each of those people, the person they are developing. Column four: anyone in generation four you know of. Now fill in what you can and leave the blanks visible. For every blank in column three, schedule one conversation in the next two weeks with the generation-two person, in which you ask a single question: who are you building, and what are you giving them? Do not accept a name without an assignment attached. Bring the completed map, blanks and all, to Lesson 7.6.
Assessment
This Week’s Commitment
Name one specific person you will begin developing — not managing — in the next fourteen days. Write down the real assignment you will give them, the decision you will let them make without you, and the date you will sit down to debrief their reasoning rather than their result.
Identity statement to carry this week: “I am not the point. The people I develop are the point, and I am free to be surpassed by them because my worth was never resting on being the best leader in the room.”
Discussion Questions
- Is there a defensible case for Paul's refusal to take John Mark — and if so, what does that say about leaders who decline to invest in someone who has already failed once?
- Where does developing a successor shade into building a copy of yourself, and what specific safeguards would catch that before it is too late?
- A leader says her bench is not ready. What evidence would distinguish an honest assessment from an identity problem wearing an honest assessment's clothes?
Reading List
- 2 Timothy 2:1-7; Luke 6:12-16; Acts 15:36-41; 2 Timothy 4:11 (NASB 1995)
- Morgan McCall, Michael Lombardo & Ann Morrison, The Lessons of Experience (1988)
- Tammy Allen, Lillian Eby, Mark Poteet, Elizabeth Lentz & Lizzette Lima, 'Career Benefits Associated with Mentoring for Proteges,' Journal of Applied Psychology 89 (2004)
- Avraham Kluger & Angelo DeNisi, 'The Effects of Feedback Interventions on Performance,' Psychological Bulletin 119 (1996)
- Michael Polanyi, The Tacit Dimension (1966)
- Jean Lave & Etienne Wenger, Situated Learning: Legitimate Peripheral Participation (1991)
- Lee Kuan Yew, From Third World to First (2000) — read critically alongside the contested record