60Immunity to Change

Robert Kegan and Lisa Laskow Lahey, Immunity to Change: How to Overcome It and Unlock the Potential in Yourself and Your Organization, Harvard Business Press, 2009

Linked to 24 chapters: supports 9 · contradicts 2 · mechanism 12 · same claim 10 · pairing 3

Kegan and Lahey open with heart patients. Told by their doctors that they will die unless they change their diet, stop smoking and start exercising, only one in seven manages it. The other six want to live. They do not lack urgency, they do not lack incentive, and the doctors have made sure they know exactly what to do. Something else holds them, and the book's argument is that the something else is not weakness of will but a second commitment, running with equal energy in the opposite direction and entirely out of sight. The two together form a system, which Kegan and Lahey call an immune system, since like the biological kind it is intelligent, it is protective, and it does harm only when it rejects material the organism needs. Behind it lies thirty years of longitudinal research on how adults' minds grow: three plateaus of mental complexity, reached in the same order by everyone who reaches them, with most adults stopping short of the second. Each plateau is not only a way of knowing the world but a way of managing an anxiety the person never feels, because the management is automatic. A change that dismantles it feels like a threat to life, which is why sincerity, incentives and willpower do not touch it. To make the system visible the authors draw a four-column map — the improvement goal, the behaviours that work against it, the hidden competing commitments that make those behaviours effective, and the big assumptions that make the competing commitments inevitable. To dismantle it they do not argue; they experiment, treating the fourth column as a set of claims about the world and designing small, safe tests to find out whether they are true.

Three plateaus in adult mental complexity

Thirty years before this book, an expert asked to plot mental complexity against age would have drawn a line rising through the teens and flat thereafter, and would have drawn it with confidence. Brain researchers on conference panels smiled politely at the two psychologists who kept reporting, from interviews conducted and reconducted with the same people over many years, that some adults went on making qualitative leaps of the kind well documented in childhood. By 2009 the hard scientists were talking about neural plasticity and the curve had been redrawn: a gentle upward slope across adulthood for a population, wide variation at any single age, and, for an individual, a staircase rather than a ramp. The time spent on each plateau lengthens, and the line grows thinner at the top, because fewer people get there.

The three adult plateaus are the socialised mind, the self-authoring mind and the self-transforming mind, and Kegan and Lahey describe them by taking one feature of organisational life — the flow of information — and looking at it through each. With a socialised mind, what I send is governed by what I believe important others want to hear. This is the mechanism under the groupthink studies, in which members withheld information they knew was decisive because they could see the leader wanted support. That effect was first read as an Asian cultural trait and Milgram's obedience experiments were designed to explain the good German; both turned out to describe people everywhere, because the source is complexity of mind and not culture. What such a person receives is equally distorted, since attention goes to imagined subtexts that may carry more weight than the message. With a self-authoring mind, I send what I judge others need to hear to advance an agenda that is mine, and I receive through a filter that admits what I asked for and what is obviously relevant to my design. That filter is an advance — focus, and the ability to separate the important from the merely urgent — and also a trap when the design is flawed or the world moves. The self-transforming mind has a filter too but is not fused with it: it can look at the filter rather than only through it, and so it solicits the information that might overturn its own design.

The whole scheme rests on one distinction, between what a mind can look at and what it can only look through, which the authors call the subject–object relationship. A small child is subject to her perceptions and says the people below the tower are tiny; an older child can look at the perception and says they look tiny. A mind becomes more complex when something that had it becomes something it has. Lahey led the development of the measure, a ninety-minute Subject-Object Interview that hands the person ten index cards marked angry, anxious, success, torn, change and five more, and then probes not what made them angry but why that would. It has been given to thousands of people and discriminates five transitional positions between any two plateaus with high interrater agreement.

The value of a higher plateau is an empirical question, and the authors answer it with small studies. Keith Eigel assessed twenty-one chief executives of industry-leading companies averaging over five billion dollars in revenue, along with promising middle managers those same chief executives nominated, and scored them separately on managing conflict, solving problems, delegating and building relationships. Complexity and competence rose together. Only four of the twenty-one chief executives were beyond self-authoring, and about half the middle managers had reached it. Two meta-analyses of several hundred people each, one using the Subject-Object Interview and one the Washington University Sentence Completion Test, agreed to the percentage point: in each, exactly 58 per cent were not yet self-authoring, and both samples were skewed towards college-educated professionals, so the figure for the general population is probably higher. Meanwhile Nathaniel Branden asks the whole workforce for self-management and self-direction, and Chris Argyris asks leaders to step outside their own ideology, see its defects and re-author it. Kegan and Lahey read both as demanding, without knowing it, a shift of one whole plateau: good soldiers asked to become sure captains, sure captains asked to become something rarer still.

Technical and adaptive challenges

Ronald Heifetz's distinction gives the book its pivot. Removing an inflamed appendix and landing an aircraft with a stuck nose wheel are technical challenges: not easy, not unimportant, but the skill is known and the route from novice to practitioner is well travelled. An adaptive challenge cannot be met by adding to the repertoire, because meeting it requires the person to grow, and Heifetz's observation, which Kegan and Lahey repeat throughout, is that the largest error leaders make is to apply technical means to adaptive problems. The test is empirical and everyone has already run it. Peter Donovan, the chief executive whose map opens the diagnostic chapter, cut out second helpings and desserts, lost ten pounds in two months and had kept them off four years later; for him that was a technical problem and a diet solved it. For most people it is not, and the average dieter regains 107 per cent of what was lost. Peter had also tried to soften his tone and release his direct reports, with temporary success and then a return of the old behaviour, plus seven per cent, by his own admission. Behaviour that returns after honest effort is behaviour doing work for someone.

Uncovering the immunity to change

Peter Donovan cofounded a financial services company in New England and, over three and a half years, acquired two competitors and took it from a hundred people writing a billion dollars of commercial mortgages to three hundred writing three and a half billion. The integration required him to be less hands-on, and he named three goals: to be more receptive to new ideas, more flexible about new definitions of roles, more open to delegating and supporting new lines of authority. Asked to inventory what he did that ran against all this, he was candid. He gave curt, closing-off responses to new ideas. He did not ask open-ended questions often enough. He communicated too often that people needed to touch back to him, and gave his own opinion when that was not what was wanted.

Working directly on that second column is what any sensible person does, and it is the technical move. Kegan and Lahey treat those behaviours instead as evidence, symptoms that will lead to the real problem if followed. The psychologist William Perry said the two things worth knowing about someone you are trying to help are what they really want and what they will do to keep from getting it; the map adds a third, why they persist. Peter's answers were a commitment to having things done his way, to experiencing himself as having direct impact, to the pride of seeing his stamp on things, and — the one that came to feel decisive — to preserving his sense of himself as the super problem solver, the one who knows best. With that column in place the entries stop being a list and become a system in equilibrium, in which the second column's behaviours are not weak failings but highly effective service to the third column's commitments. Peter described the result as a picture of himself with "one foot on the gas and one foot on the brake."

The immune metaphor is used deliberately and in both directions. An immune system is not a fault; it is an intelligent force that acts to protect and often to save a life, and so is every immune system these maps reveal. A member of Peter's senior staff, looking at his chief executive's map, said approvingly that the stubbornness in the picture had a great deal to do with the success of the company and the size of his own house. It becomes dangerous only when it rejects material the organism needs, and then it is not malicious but mistaken: it does not understand that it must alter its code. What the map shows is not an insincere commitment behind a sincere one. Change fails, the authors write, because we mean both things; the heart patient is not insincere about wanting to live as he reaches for another cigarette.

The third column is not predictable from the second. Ron Halpern, Peter's chief operating officer and his friend of many years, made hard decisions daily except within the executive staff, where he was not direct, overconsulted and tried to please everyone; his third column proved to be a commitment to being well liked and to preserving his closeness with Peter, though another person with the same first two columns might be guarding against blame, or against being envied. The authors make the point with weight loss, where the first two columns are nearly universal and the third columns differ entirely. One man found that his overeating treated emptiness rather than hunger; another that his aunts' repeatedly offered manicotti was not an extra portion but love; a woman who lost twenty pounds and regained them each time found she was committed to keeping her relationships desexualised. Three adaptive challenges, and a diet for none of them.

Getting to a third column requires the worry box, an empty square at the head of the column. Take each second-column behaviour, imagine doing the opposite, and name not the thought but the feeling. The first answer is usually a cover. Fred, an executive whose goal was to listen better, first reported boredom and impatience; boredom is disengagement, and we disengage for a reason. Pressed, he found that if he stayed present with his teenagers he would meet rolled eyes and a disdain close to unbearable, and that if he stayed present with his wife he would feel helpless about matters outside his control. Six fears went into his worry box, and each converted into a commitment to keeping that thing from happening. A commitment to not looking stupid makes withdrawal not merely understandable but excellent work.

The fourth column holds the big assumptions, called big not because they are grand but because they are not held as assumptions at all. They are taken as facts about how the world is, and while they may be true, the question cannot be asked while they sit behind the eyes rather than in front of them. Bringing one from subject to object is the motion by which complexity increases anywhere, which is why the map does three jobs at once: it shows how a person prevents the change they want, how a persistent anxiety is being managed, and what must hold if they are to keep knowing the world as they do. Peter's third column reads like a self-authoring mind defending its authorship; Ron's, whose deepest satisfactions come from alignment, reads like a socialised one. Each plateau has its own world of dread.

Seventy-five years of developmental research, the authors say, gives one answer to what produces greater complexity: optimal conflict, a persistent frustration perfectly designed to make us feel the limits of our current way of knowing, in some sphere of living we care about, with enough support that we are neither overwhelmed by it nor able to escape it. Drawing a map manufactures exactly that. It takes a contradiction the person is — which is why they systematically cannot do what they intend — and converts it into a contradiction they have, and can work on.

The first column decides everything downstream, a point Peter understood before his consultants did. Their first session with his executive committee was three absorbing hours that, a year later, had changed nothing, because people had chosen their goals on the spot. The second attempt began weeks earlier, with every member asking the person they reported to, several peers and at least one direct report a single question: what is the one thing it would be most valuable for me to get better at? The phrase one-big-thing entered the company's vocabulary, and Peter added the requirement that a goal also pass the spousal test, which he tried himself and failed, being told after twenty-three years of marriage that his little problem with control was a large one. At the offsite that followed he wanted all eighteen members to share their maps the next morning; a vice president said his people were not ready, then wondered aloud whether what was happening right then was an example of Peter's own big thing. Peter backed off, and thirty days later every member shared voluntarily.

The collective immunity

Groups have immune systems too, and the same four columns draw them. A humanities department at a leading research university had not tenured anyone from its junior ranks in eleven years and could not work out why. Its members mentored, told the junior faculty that publishing mattered, and gave semester leaves; they also loaded them with advisees and committees, gave them heavy teaching, and signalled that everything else mattered too. Asked what they would be most worried about if they did the opposite, they laughed as they answered: if the junior faculty did not take the advisees, the teaching and the governance work, the senior faculty would have to. One member then told a story against himself. He had coached a promising young colleague on how to decline committee invitations, and two weeks later telephoned her, wearing his admissions chairman's hat, to ask whether she could find the time.

The collective third column is where groups get stuck, and the school district shows why. The leadership team of a district in Southern California, serving a student body more than 80 per cent Latino with a professional staff more than 80 per cent white, set out to accelerate the achievement of its English language learners and admitted in the second column that it did not hold high expectations for them. Its first third column named the extra work that higher expectations would demand of an already stretched group. The map was technically correct, produced no energy, and the session adjourned. The next morning an agitated assistant superintendent, himself Latino, said the group was not telling itself the truth: what belonged there was a commitment to preserving a pobrecito culture, protective and sympathetic, whose logic ran that these children already carry so much that it would be cruel to add rigorous expectations. He judged that no white administrator could raise it without fearing to look racist, and that if he could not raise it nobody could. One member said afterwards that what opened it up was the idea that low expectations could come not from disregard but from love.

Where groups carried the work through to tests, the authors report outcomes. Twenty partners of a strategy consulting firm, holding an assumption that entrepreneurship and collaboration are an either-or, tested it by identifying ten new projects to be hunted in packs. An outpatient clinic in an academic medical centre, where physicians were a soft touch for patients seeking narcotics and nurses were undermined when they held the line, mapped doctors and nurses separately and found the same structure with different audiences: the doctors feared the patients' reactions, the nurse feared the doctors'. Five months later, fourteen patients had been discharged from the practice for contract violations against none in the previous two years, narcotic contracts had risen from 30 to 65 per cent of those requesting refills, and every nurse on the prescription line reported feeling supported in enforcement, where before almost none had. At an American medical school, Constance Bowe and colleagues found that the faculty ranked small-group discussion above lectures for every learning objective but two, while lectures made up 65 per cent of the preclinical programme and small groups 12 per cent. That contradiction, produced by the faculty about themselves rather than imposed on them, became the second column, and the reform then proceeded as pilots designed to test assumed limitations of the students, the faculty and the institution. More than eighty faculty from seventeen departments volunteered in the first year.

David and Cathy

David, in his mid-thirties, had been promoted to general manager of a worldwide engineering company six months before the authors met him, and for the first time in his career felt overwhelmed. His goal was to focus on a few critical things, and he knew delegation was part of it. His second column was short. He let new opportunities lengthen his list, accepted more tasks and paid for them with sleep, family and hobbies, and did not ask people to help him. His third column made those behaviours excellent: they kept him independent, capable of anything and selfless. Rereading the map days later he saw what tied them together and added it himself — a commitment to being loyal to his blue-collar roots, and beneath it a definition of delegating as getting other people to do your work, which can only ever look like laziness. His summary assumption was that leadership without doing is overhead and worthless, that he would be walking away from his roots if he were not doing the work himself, and that he would be selfish, lazy, spoiled and without self-respect.

What followed was not a delegation technique. David told his team what he was working on, which gave them permission to tell him when they saw him doing work they could do better. He filled in a continuum of progress — first steps, indicators of significant progress, what full success would look like — and sent a short anonymous survey to six or eight witnesses who would be in a position to notice real change, which served at once as a check that the goal was the right one and as a baseline. He asked all thirty-six people in his group what five things they would want to know if he gave them an assignment, and the answers differed completely: one colleague, told to produce ten phones by tomorrow, did not deliver and had begun to look lazy to him, but told why and for what could do ten by three o'clock. Reading his survey results he saw something worse about himself, that by giving deliberately thin directions he had been preparing people for failure and setting the stage for his own rescue — as clear an instance as the book offers of a third-column commitment moving from subject to object. He then audited six months of his calendar against the three activities he now considered his real work, and found a quarter of his time on strategy, a quarter on people, a quarter on tasks and a quarter wasted. His revised assumption was that it is not that he needs to do it, but that he needs to know it, and he described the change as moving from tradesman to architect. The line that marks the turn he offered ruefully: when you give people a chance to shine, they do it better, so his assumption that he would do it better was not correct.

Cathy, a marketer at one of the largest pharmaceutical companies, wanted to contain her upset to the situation that caused it rather than letting it spill into everything else. Her colleagues confirmed the goal and insisted, in the same surveys, that her emotional energy was an asset that must survive the fix. Her third column was a commitment to never letting herself or a team down, whatever it cost, and her fourth held the arithmetic that made it impossible: a good team member gives 110 per cent, a good Cathy gives 150, and it is worth the risk of burning out rather than give less. Yoga, breathing and exercise would not help her, the authors note, because she would do those at 150 per cent too.

Her test arrived unplanned and was the opposite of the safe, modest first experiment the book advises. Before a presentation in Houston to executives who had rejected her team's previous proposal, Cathy worked two days almost without sleep, and on the morning of the meeting she passed out. In the ambulance, with no memory of it afterwards, she repeated that they were going to be late; she was treated for exhaustion and discharged two days later. A colleague presented in her place, the proposal was accepted, and the executives told her afterwards that the structure she had built had decided the matter. What makes this a test rather than an event is what she did with it: she had done precisely what her assumptions forbade, and the consequences contradicted them. Asked whether there was anything her boss could have said that would have convinced her of his regard, she realised there was nothing, and found herself back at a rejection from medical school ten years earlier, after years given to nothing else, which she had never spoken about and never grieved. Her largest assumption never made it into her fourth column at all: that there is a flaw in her, and that she must run her life as she does to keep the world from finding it. Both cases end in the same shape, with what is accomplished going well beyond the goal that started it.

Designing, running and interpreting a test

The work takes about thirty minutes a week and most people notice change in about twelve weeks. It opens by honing the map and surveying the witnesses; its middle is a continuum of progress, self-observation of the assumptions in action, a biography of each assumption asking when it started and whether it is still valid, and then repeated tests; it closes with a follow-up survey to the same respondents and an exercise on hooks and releases.

A test must first be testable. An assumption containing the words die, be fired or have a nervous breakdown is not ready; the person backs up the chain to an earlier link — if I disagree with my boss I will be fired becomes if I say this, my boss will be angry. The design must then be safe, so that a bad outcome leaves the person able to run another test another day, and modest, a single step past the warning sign rather than a move to somewhere miles beyond it. It must be actionable within a week or so, ideally inside an ordinary day rather than requiring a special occasion. And it must be run from a research stance rather than a self-improvement stance: the purpose is information about the assumption, not getting better. The check on this last point is decisive — if you cannot name in advance any result that would cast doubt on your assumption, you do not have a test.

Sue, chief of staff of a large social services agency, turned an assumption that she had no value unless she was accepted into something she could act on: if I say no, others will see me as cold and uncaring. She named the situation that most reliably triggered it, chose the two colleagues with whom the risk was acceptable, rehearsed what she would say — barking a refusal and declining warmly are not the same test — and decided in advance what she would watch. When the moment came she declined to get involved, did it without feeling bad, and was telephoned later by the colleague, who apologised for dragging her in. Once saying no proved survivable she found the question underneath it: could she say what she actually thought?

Interpretation is where the discipline lies. The point is rarely to reject a big assumption outright, since one is seldom completely right or wrong; the trouble is that it is overused far beyond its proper scope. A good test sharpens its contours until the person has a data-based account of when, where and with whom it applies, and even modest changes of that kind can overturn an immunity. No single experiment settles anything, and a test can be complete even when the action fails, provided it produced data. Tests are therefore run in series, each a version of the last with different players, circumstances or risk, until the assumption no longer has force — at which point the self-protective commitment is no longer needed and the obstructive behaviours stop being generated.

Three things predict who gets through. The first is gut: the goal must be more than important, and the authors report people who rated a goal four or five out of five and then quietly downgraded it the moment they saw what accomplishing it would cost their self-protection. What supplies the necessary fire is usually the sense that not changing is putting someone at risk, as it did for Fred when his daughter's therapist suggested her behaviour might be connected to his unavailability. The second is head and heart together, because a mindset is also an anxiety-management system and no amount of thinking alone will move it. The third is hand: Kant said perception without conception is blind, and the authors add that conception without inception is paralysis. Reflection without action is as unproductive as action without reflection. The collective version follows the same steps and the same criteria, and what makes a group's map bite is shown by research university librarians who wanted a real voice in governance and feared that, given one, they would be exposed as frauds. Their first third-column entry, that they were committed to sticking with their strengths, drained all of that away; rewritten as a commitment to taking no action that might expose them as frauds, the map worked.

Since 2009

The four-column map has outlived the argument that produced it. It became a standard instrument in executive coaching, taught through the practice Kegan and Lahey founded, and the vocabulary — competing commitments, the big assumption, the immunity map — passed into general use in leadership development, often detached from the developmental theory underneath. The authors extended the claim in 2016 with An Everyone Culture, written with Matthew Miller, Andy Fleming and Deborah Helsing, which studied three organisations that had built continuous development into daily operations and argued that a whole company can be run this way. That book drew the sharper criticism of the two, since the cultures it admired were exacting and, in one prominent case, widely reported as harsh; the line between a setting that supplies optimal conflict and one that supplies ordinary pressure is easier to draw in a diagram than in a workplace.

The developmental theory has aged less comfortably than the method. Constructive-developmental stages remain contested: whether adult growth is genuinely discontinuous, whether the ordering is as invariant as the longitudinal interviews suggest, and whether a trained rater scoring an interview measures a structure of mind or a style of speaking about oneself. The evidence tying complexity to performance here is thin by the standards now expected — twenty-one chief executives in one dissertation, eight leaders in another, sixty-four consultants in a third — and the distribution figures come, as the authors say plainly, from samples of college-educated professionals. The heart-patients statistic that opens the book carries no citation at all, and versions of it circulate with the ratio varying from one in seven to one in nine. Heifetz's technical-adaptive distinction, by contrast, has held, and was itself extended the same year in The Practice of Adaptive Leadership.

What has lasted is the mechanism and the move. The mechanism is the observation that resistance to change is usually not resistance but a second commitment doing excellent work, and that the way to reach it is through the fear that surfaces when you imagine doing the opposite of what you do. The move is the conversion of a belief into a claim and then into an experiment — small, safe, run to get information rather than to improve, interpreted for what it says about the belief and not about the performance. That is a practice anyone can adopt without accepting the plateaus, and most of those who use the map now do exactly that. The authors' own warning has also held up, and reads more sharply now than it did: "We used technical learning designs and expected them to deliver adaptive results."