Power Without a Diagnosis

Popular talk about power loves medical labels: the boss is a narcissist, the politician a psychopath, the manager a sociopath. A diagnosis handed out by job title helps little; observable behaviour and the design of authority say more.

In 1963 Stanley Milgram published the results of an experiment in which participants, at the researcher’s insistence, kept pressing levers that supposedly delivered ever stronger electric shocks. The work became a symbol of obedience to authority — and at the same time a target of serious criticism for its ethics, the pressure on participants and the simplified retellings that followed.

The reliable conclusion is more modest than the popular one. An ordinary person can go a long way inside a procedure that looks legitimate, if responsibility is distributed, the action is broken into small steps, and authority keeps repeating that everything has been taken care of. So it is worth checking not only the boss’s character, but the design of the process that makes refusal more and more expensive.

It does not follow that power is nothing but belief.

Behind it there may be:

• the right to fire;
• access to money;
• control of information;
• the police;
• the army;
• property;
• a professional licence;
• the ability to shut a person out of an important system.

A boss has power not only because a subordinate has inwardly agreed to play a role. Losing a job can have entirely material consequences.

Any power, however, is limited by dependencies. A manager needs staff, a state needs enforcers and taxes, an owner needs customers and working infrastructure.

This creates room for choice, but it does not guarantee that the sides are equal.

A person can indeed often choose a boss — by changing job, profession or country. The price of that choice varies. For one it equals a few interviews, for another the loss of housing, papers and health insurance.

A good manager does not give up power. He makes it legible and bounded:

• explains decisions;
• allows objections;
• does not demand personal loyalty;
• takes responsibility;
• shares information;
• builds the independence of the staff;
• does not turn dependence into a source of pleasure.

It is worth choosing people you can respect for something other than their post. But a decent person also gives unpleasant orders at times. Management is not obliged to be friendship.

The best criterion is not the absence of power, but what happens when you disagree.

So in judging power it is more useful to look at authority, incentives and observable treatment of dissent than to hand its holder a psychological diagnosis.

But power creates several persistent risks: less honest feedback, more people with an interest in agreeing, the option to pass consequences on, the habit of deciding for others, and less contact with those who live with the results of those decisions.

A person may grow harsher not out of sadism, but because he works with numbers rather than faces, constantly makes unpleasant decisions, or has not encountered safe disagreement in a long time.

For a subordinate, what matters is not guessing the boss’s disorder but assessing observable behaviour:

• does he punish bad news;
• does he change agreements;
• does he humiliate people;
• does he provoke conflict;
• does he demand personal loyalty;
• does he take the credit and hand out the blame;
• is it safe to object;
• can you leave.

Shouting often signals that other means of influence have run out, but sometimes it serves as a deliberate instrument of intimidation. In both cases volume adds no arguments.

A good diagnostic sign of character is how someone treats a person from whom nothing is needed: a waiter, a cleaner, an intern, a driver. Politeness only up the hierarchy is called calculation, not good manners.

If an environment systematically humiliates and punishes, the main task is not to out-manoeuvre its author but to preserve documents, contacts, professional reputation and a way out.

The sign can be tested against deadlines. In a department where bad news is punished, every project is on schedule until the last week and then slips all at once. Nobody lied in a big way: each reported slightly more optimistically than he knew, because bringing a problem first is more expensive than bringing it together with everyone else. And there is nobody to diagnose: the mechanism is assembled out of normal people, each of whom rounded a day or two in his own favour.

So the quality of a manager shows not in calm times but in who learns about a missed deadline first. If the boss finds out last, he is no longer managing the work, only the reports about it.