Domains of Incompetence
In physics, models are built on top of of axioms and first principles. An axiom is something taken as true. A first principle is an axiom or set of axioms that cannot be further simplified.
An example of an axiom is the definition of a real number. An example of a first principle is the definition of a derivative:
In physics, both of these concepts can be directly measured/observed. But, using the same lens to view more qualitative principles runs the risk of establishing false premises. I view leadership through that lens; one axiom I’ve developed is that good leaders are stable under pressure, and annother is that stability comes with mental “normality” (some call them “homoclites”). The resulting first principle is that mentally stable people are most suited to make decisions.
That’s a cornerstone of my beliefs. It’s the reason I scorn Penn’s alcoholics, social climbers, and mental trainwrecks as wastes of an education. In my head, abysmal decision making – even among young adults – should be enough to disqualify certain people from any future power or influence.
That principle can be disproven through observation. An uncomfortable fact has always been that many of those peers will become the most successful: the CEOs, Prime Ministers, and Nobel Laureates. My mental model finds ways to deal with the error terms: it alleges that many succeed in spite of these issues. Perhaps those traits correlate with economic or social privilege, which outweighs poor decision making. Perhaps there were so few radically successful people that outliers must pop up from time to time. Or, most in line with my beliefs, perhaps our society is bad at picking who becomes successful.
But those explanations seem insufficient. If my beliefs are true, at best, the average unusually successful person would be mentally normal and the tail ends would be the depressed and manic. In real life, both tails are highly overrepresented. Why?
Recently, I read Nassir Ghaemi’s A First-Rate Madness, which champions an unconventional thesis: that successful leaders (Lincoln, FDR, Kennedy, MLK, Gandhi, Churchill) succeed because of underlying mental illnesses. Their abnormalities in times of peace became necessary in times of crisis. On the flip side, leaders like George McClellan, Neville Chamberlain, and George W. Bush made disasterous mistakes due to the absence of those characteristics.
Psychoanalyzing historical figures is a controversial practice at best, and Ghaemi’s conclusions are highly disputed. His arguments are best read firsthand, but one of the most fascinating arguments is that depressed people view the world more realistically than their homoclite counterparts. To be “mentally healthy,” you need to overestimate the degree of control you have on your life – and that has some dramatic consequences for leaders. When making the wrong decisions, leaders that haven’t experienced sufficient negative feedback will discount negative feedback and persist. The consequences of this can be severe (excerpt from A First-Rate Madness):
When we started to lose, we tried harder, consistent with [the] dictum that fanaticism consists of redoubling your efforts after losing sight of your goals.
[George] Bush’s defenders will offer more complex rationalizations for the invasions of Afghanistan and Iraq, some of which might be valid. But the invasion of Iraq clearly was based on false claims. Rather than admit error, as Kennedy had in the Bay of Pigs, and withdraw, Bush did what Kennedy refused to do: send in even more troops and stay even longer.
This is quite normal. In psychological terms, Bush can’t be faulted for thinking this way. Most of us think and act similarly. Most people have a hard time admitting error, apologizing, changing our minds. It takes more than a typical amount of self-awareness to realize that one is wrong and to admit it.
I know almost nothing about psychology, so I can’t speak to his conclusions. Still, the book made me question my definition of an effective leader as one that makes the right decisions whenever prompted. In reality, even the best leaders have incredibly narrow domains of expertise. When confronted with unfamiliar situations, the humble, effective response is recognizing ineptitude and deferring authority. When receiving negative feedback, they will switch course.
But, no leader will be praised for taking that course. Deferring authority or changing decisions are (incorrectly) perceived as weakness. Consequently, many (like Bush) would rather preserve their egos. Some (like McClellan) were so unused to failure that they didn’t even consider that they were taking the wrong approach. In contrast, General Sherman’s March to the Sea proved unconventional and incredibly effective, just as FDR’s New Deal was ambitious beyond belief but necessary. My paradigm for evaluating who “deserves” to lead based off stability, pedigree, and past achievement is wrong.
That leads me to a second jarring realization: I’m bad at evaluating my strengths. By college, most people have made few decisions of consequence. If a decision has had good outcomes, it could be the result of excellent judgment – or luck. My model of thinking always tries to maximize my agency in a situation. When an outcome is desirable, I credit myself too much; when it isn’t, I’m my biggest critic. Trained off such a small sample size and the tenuous assumption of causation, the resulting model of my strengths and weaknesses is probably deeply inaccurate.
Similarly, General McClellan thought that his thesis on the Napoleonic Wars, position at West Point, and the initial trust of his superiors meant his strategies were foolproof. As it turns out, those were the wrong metrics to optimize for – conventional wisdom was wrong. McClellan misread his own expertise. Sherman, without as inspiring of a track record, found the unconventional solution – evade major conflict, and just burn the South’s infrastructure to the ground to win the war.
If outcomes don’t always validate our actions, what can? An alternate approach might be how some soccer clubs optimize for expected goals (xG) instead of goals. The argument is that optimizing for goals and rewarding only goals doesn’t make sense because of the high variance (i.e. the goalie). Many teams have instead started rewarding those who are in positions to score. In the long term, optimizing for xG, and not just wins, will lead to more wins.
Measuring the outcomes of decisions might be similar to measuring goal scored: the metric is subject to randomness. Judging for the objective quality of decisions is, of course, harder than calculating xG. I’ve spent my life calibrating a binary good/bad outcome model, which has made me overconfident where I only have a narrow edge.
I didn’t make a good crisis leader. Last August, I wasn’t able to persist/pivot fast enough. I couldn’t keep morale on my team alive. Since then, I’ve put myself in a self-exile. My days revolve between reading, academic work, and a strict exercise regiment. I’ve tried regulating my emotions to the maximum possible extent. In my head, I’ll eventually be “perfect.” I’ve refused to start anything new until I can be the picture of ideal leadership; it would be irresponsible.
Even if I got there, Ghaemi would argue that there are situations where “ideal leadership” is a structural disadvantage. The better outcome would come from ceding authority to the more capable, or being more willing to course correct. The bizarre conclusion of this thinking is that, no matter how capable or educated, a better leader is less steadfast in his beliefs.
An even more bitter pill is that the right person to take authority might be a type I don’t respect: someone volatile, with poor virtue and values, less focused on orderly systems, and less knowledgable. In my head, I easily acknowledge that the dyslexic are often more creative, or that the less empathetic make better soldiers. Still, I struggle with the logical extension: that those with different values can make better leaders than me. Even deep domain expertise wouldn’t mean my decision making was superior in every situation
In my next time of crisis leadership, I hope I have the humility to acknowledge my domains of incompetence – and prioritize outcomes over my ego.